ill... 11:21. 3}: .1: 3'1, I!!! \\ .8 . , it II». x... Al: 2 L .7. . 1 t. i I» 1 . a at." . £.l~..F .. .3: .. v‘I;‘ari ll- :tn 15.). . .2: an .n . . : putxhl. 1... )33353. .4}: in». xx» {no . .g‘ ”7." ' E I Y $3' $533 " vn mug-31!! 3 lllllllllllllWWlllHlWHllllHlHll\Hllllllllll 3 1293 014214 This is to certify that the dissertation entitled Structural Determinants of Discount Department Store Locations in the Central Cities of the Top 50 Metropolitan Areas presented by John Terrence Farris has been accepted towards fulfillment of the requirements for Ph.D. degree in Social Science Date W MS U is an Affirmative Action/Equal Opportunity Institution 0-12771 - v-——.L_.——__-i LIBRARY Mlchlgan State University PLACE IN RETURN BOX to remove this checkout from your record. To AVOID FINES rotum on or bdoro date duo. DATE DUE DATE DUE DATE DUE MSU to An Affinnnttvo AotchEqml Opportunity Ila-titular: Wt STRUCTURAL DETERMINANTS OF DISCOUNT DEPARTMENT STORE LOCATIONS IN THE CENTRAL CITIES OF THE TOP 50 METROPOLITAN AREAS By John Terrence Farris A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY College of Social Science Urban and Regional Planning Program 1996 ABSTRACT STRUCTURAL DETERMINANTS OF DISCOUNT DEPARTMENT STORE LOCATIONS IN THE CENTRAL CTTIES OF THE TOP 50 METROPOLITAN AREAS BY John Terrence Farris Since the 19608, discount store retailing has become a predominant mode of shopping throughout the United States, with over 11,000 stores. Full-line discount department stores include the Big Three-Wal-Mart, Kmart, and Target-and many regional firms. Individual store lifecycles are evolving quickly with larger and broader formats including grocery sales as the new wave; early stores of 40,000-80,000 square feet are rapidly becoming obsolete, abandoned, or replaced by units as large as 200,000 square feet. Most stores have located in the suburbs and rural hinterlands. Large central city planners believe that retailing is an important component of the city's future through tax creation, job opportunities, and shopping opportunities for city residents. Little research has been given at the variables affecting discount store locations in central cities compared to their related suburban areas--most research has been done at the metropolitan level based on sales or number of stores for general merchandise, and not by square footage. Discount department store space for 1992 in the Top 50 Primary Metropolitan Statistical Areas is evaluated at the central city and suburban level, representing 2,435 stores. Median square feet per eapita in central cities is 1.2 compared to 2.6 in suburbia. A proposed multiple-regression model is tested hypothesizing that each central city's capture of fair share space (based on gross income) is dependent on seven variables pertaining to the amount of housing built in the last decade, density, race, income, violent crime, and percent of local revenue generated by sales tax. Variables represent ease of land assembly and investment opportunity with the general hypothesis that the central cities getting their fair share capture are those most closely resembling their counterpart suburban areas. The top fifty central cities range from zero to 190% capture, with a median of 64%. Only the amount of housing built during the last decade was deemed significant, representing both lifecycle and case of land assembly factors. Central cities should evaluate their retail portfolio for potential opportunities, especially if they have older, small discount stores subject to restructuring. iv To my Father, whose enduring planning and public service career since the thirties has instilled me with the desire to continue assisting the betterment of the nation's communities. ACKNOWLEDGMENTS I want to thank Michigan State University for providing five years of inspiration through my Masters and Ph.D. programsuat two significantly different junctures of my life. I truly enjoyed the opportunity to retum after a seventeen ycar planning career for a long educational sabbatical including the opportunity to teach. Certainly, I want to thank the many professors and staff in geography, public administration, and urban and regional planning. Special thanks to my dissertation committee-Carl Goldschrnidt, Assefa Mehretu, Ric Hula-~with Roger Hamlin as Chair. Everyone was very supportive and kept me appropriately focused. I am proud to be a Spartan. Finally, I want to thank my father, family, and friends in St. Louis, who gave me positive support on long-distance calls and brief vacations. I sometimes miss St. Louis, but I have made the right decision to enter academia. Throughout my remaining career, I will have the fortunate opportunity to see my students positively affecting the quality of life for many people and communities. vi TABLE OF CONTENTS LIST OF TABLES ix Chapter 1. INTRODUCTION 1 Purpose of Study and How This Proposal Advances Theory Research Outline 2. LITERATURE REVI EW—PART I DISCOUNT STORE AND SHOPPING CENTER STATUS 7 Historical Evolution of Discount Stores The Discount Industry In 1992 Discount Market Share Stateofthelndustryandthe BigThree Kmart Stores Wal-Mart Stores Target Stores Urban Market Entry Trends SupercentersuFood and General Merchandise Combined Shopping Center Development Status 3. LITERATURE REVIEW--PART II THE SPATIAL AND CENTRAL CITY DEVELOPMHVT PERSPECTIVES 51 Spatial Investment Strategy The Risks of Investment Factors Affecting Retail Development Trends Central Cities vs. Suburban Trends Shopping Demographics/Needs Life Cycle for Development and Trade Areaso-The Central City Dilemma Why Central Cities Need Discount Stores--The Pro Growth Perspective vii Table of Contents (cont'd.) Local Government Revenue Sources Barriers to Redevelopment Urban Politics The Elasticity of a City 4. LTTERATURE REVIEW--PART III MODEL THEORETICAL FOUNDATION 96 Optimum Network/Portfolio Macro Structural Determinants Selected Retail Location Variables 5. PROPOSED MODEL, HYPOTHESES, AND METHODOLOGY 116 Overview The Dependent Variable Proposed Model Hypotheses Methodological Overview Delineation of Market Study Areas for Data Collection Market Disaggregation and Assumptions Data Collection Sources Data Analysis and Hypothesis Testing Procedures 6. TEST EVALUATION AND DISCUSSION 134 Test Results Studentized Residual—Single Independent Variable Model Insignificant Independent Variables Descriptive Data Caveats General Merchandise Sales versus CAPTRLQ Degree of Elasticity Table of Contents (cont'd.) 7. SUMMARY AND CONCLUSIONS Retail Difficulties in Heterogeneous City Markets Market Research Difficulties Initial Steps for City Planners to Attract Retail Developers Future Research Concluding Remarks GLOSSARY OF TECHNICAL TERMS APPENDD( A1. Top 54 Discount Store Chains in Sales, 1992 A2. Select Data Sorted by Dependent Variable REFERENCES CITED REFERENCES 150 163 166 169 172 180 LIST OF TABLES Table 10. ll. 12. 13. 14. 15. l6. l7. . Sales And Number of Discount Stores By Store Size, 1992 Discount Store Profile, 1992 . Sales And Ownership Of Discount Stores By Company Size, 1992 Sales Share By Major Departments, 1992 Big 3 Discount Department Store Trends, 1993 Monthly Visitation and Spending by Store Type Percentage of Stores Shopped in by us. Consumers in 1991 Kmart's Revised Development Program , 1990-1996 Nationwide Shopping Centers, 1993 Shopping Center Productivity General Revenue For All Municipalities, 199091 City Government Finance By Population Size, 1990-91 (Local Tax Sourcea As A Percentage Of Municipal Tax Revenue) Potential Land Assembly Cost For 200,000 Sq. Ft. Shopping Center In Built- Up Urban Area Revised Model of Market Attractiveness Attractiveness of Subareas Selected Retail Location Variables Independent Variables l4 14 15 15 18 20 22 45 A E 111 113 119 List of Tables (cont'd.) 18. Proposed Model Regression 19. Single Variable Model Regression 20. Single Variable Model Residuals, Actual Y, Predicted Y 21. Model Independent Variables Sorted by Dependent Variable A.1 Top 54 Discount Store Chains in Sales, 1992 A.2 Select Data Sorted by Dependent Variable 136 137 138 139 166 169 CHAPTER 1 INTRODUCTION Spatial forms of settlement and economic activities are created by economic, political, social, cultural, and geographic processes. And economic geography focuses on the interrelationship of production, distribution, and consumption of wealth. This research emphasizes the distribution process, especially through the discount store industry, for the central cities of the top fifty metropolitan areas in the United States. Central cities are attempting to strengthen their retail tax base. That base has eroded substantially since the 19503, primarily because of continuing suburbanization. The residential population shift from the central cities has spurred the growth of both suburban employment and retail development. Shopping centers have become the primary generator of retail sales in the US. capturing $814 billion, or 56% of the non-automotive retail market. In the last forty years, approximately 40,000 shopping centers have been developed compared to only 100 existing in 1950 (O'Neill 1991). Much of this development has been in the suburbs following the residential migration from the central cities. Shopping centers have located where population and concomitant higher income groups have located in order to penetrate new and expanding markets. Furthermore, stores have located in the suburbs because of excellent transportation, capital availability, land availability and zoning, size and design considerations, and local suburban governments encouraging development because of the substantial tax benefits (Craig, Ghosh, and McLafferty 1984). Suburbanization accelerated after World War II facilitated by automobile ownership, rising incomes, and govemmental incentives for homeownership. With the new interstate highways, suburbs spread into the inner interstices between the fingerlike radii of the 1 2 railroad or mass transit lines in major cities. Suburbs of increasing diversity clouded the eariier socio—economic symmetry of the radii and bands (Berry, Parr, et. al 1988). Whereas federal policies have encouraged suburban growth (i.e. transportation, water and sewer, housing, and tax policies), central cities have had more difficulty attracting major developments. Central cities have been hurt by this rapid suburban growth. They have lost a significant portion of their tax base, job opportunities, and modern retail development. Lack of retailing negatively influences the retention and attraction of residents and the preservation of neighborhoods, thus creating a vicious circle. Restructuring of the retail industry and the rapid growth of new discount store formats and shopping center configurations have serious ramifications for older areas, especially central cities. Approximately one discount store exists for every 10,000 households. Discount department stores are growing in scale and in their ability to capture an increasing portion of retail sales; furthermore, whereas discounters used to compete primarily with traditional department stores, they now aggressively compete with food (the major retail industry), apparel, drugs, cosmetics, and other retail sectors through increasingly scrambled merchandising (i.e. blending of different lines of trade in large store configurations). Central city stores of all types are susceptible to competition from these expanding discount formats. To a great degree, discount store location strategy depends on the life cycle of discount store formats, their locations, and of the trade areas they are dependent upon for consumers. So, the discounters location strategy for investment and operations is affected directly by the stability and growth of the neighborhoods served. And neighborhoods are also dependent upon convenient, quality retailing as part of the I‘bundle of services“ offered to potential residents and property owners locating or investing in an area. Both neighborhoods and stores attempt to modernize and avoid functional and economic obsolescence. As a result, some discounters are now shifting to both micro-marketing and an increasing level of individual store control. This is happening 3 in order to meet the market segmentation occurring throughout society, especially in very heterogeneous central city environs. This shift may provide new opportunities for discount store development in central cities, especially given the perceived or real saturation occurring in some metropolitan suburban areas. Retailing, and specifically the discount department store industry, has become a major generator of taxes and jobs. Consumers have indicated their preferences for large-scale, value-oriented retail distribution centers. This research is not intended to debate the validity of the format and the overall implications to society of large-scale retailing in automobile- oriented configurations. This analysis assumes the political economy as it generally operates nationally. Recommendations for future research beyond the existing framework, such as metropolitan consolidation or land value taxation policies to enhance city development, will be included. Every city needs to consider benefit-cost analysis for any development proposals that might come forward recognizing its own market conditions, financial capacity, legal and administrative aspects, economic base and social impacts, and citizen and legislative processes. Purpose of Study and How This Prom Advances Theog The market has reacted favorably to discount store development. This research analyzes how central cities have performed within their metropolitan areas and tests an hypothesis regarding the structural determinants affecting the existing full-line discount store portfolio in the central cities of the top fifty metropolitan areas. Full-line discount department stores include Wal-Mart, Kmart, Target, etc.; not stores like Office Depot, The Sports Authority, or Toys 'R' Us. The evaluation is cross-sectional based on 1992 store location and square footage information; given the rapid growth of the industry and especially the recent major restructuring plans announced for eliminating many of Kmart's older units, this analysis is static yet may provide some clues for possible plans and how communities might be affected. It should be noted that future store plans are considered proprietary and generally 4 unavailable; but cities should be forewarned that any older, small units in either the discount or food store industry make them highly susceptible to alteration, elimination, or competition. Much of the literature focuses on micro-retail analysis from the shopping center or store locational perspective. Scholars have continued to refine the Reilly and Huff retail gravity models for consumer patronage. While many of the variables used in evaluating site locations are relevant, the specificity of factors for individual sites is beyond the realm of this analysis and would be appropriate for further research into individual community market evaluations. Little research has been done on decisions to enter large geographic markets (Ingene 1984). This research evaluates a specific industrial classification, discount department stores, that has not been analyzed from either a square footage or city versus suburban competitive context. Geographical market orientations have tended to use census sales data by line of trade and by number of stores. Yet, square footage is vital in analysis of retailing today. For example, Kmart has stores ranging from 40,000 square feet to 175,000 square feet; certainly results from studying the numbers of stores can be skewed by such ranges. Furthermore, most analyses are by general line of trade at a low-level of Standard Industrial Classification such as general merchandise sales or department store sales. This research provides a detailed exploration of the f ull-line discount department store industry with actual square footage in 1992. Little analysis of central city retailing exists other than total retail sales by standard industrial classification. Most analyses emphasize differences among metropolitan statistical areas, either to eliminate the difficulties of analyzing leakage of sales on an intra- community basis or to focus on a broader basis of evaluation nationally. This research looks at the cross-sectional perspective of the central city in its metropolitan context. The dependent variable in the proposed model incorporates the leakage factor. 5 Many central city studies do not consider the market position of each city within its own metropolitan market context with suburbia; this analysis attempts to incorporate such variables. Furthermore, this analysis shows the diversity of central cities throughout the country with many continuing to grow, albeit virtually all cities have more negative traits than their suburbs. Some studies briefly identify land assembly as a problem, but this analysis specifically provides an example of the required land for retail development, the difficulties of assembly regarding acquisition and relocation, and the cost ramifications that suggest the role of the government sector to eliminate negative extemalities. A practical element of the research is the description of the discount store and shopping center industries based on a thorough review of the trade journal literature. This provides the existing and projected scale of development, including important restructuring trends in these industries. Such trends as multi-level discount stores, supercenters incorporating food, and power shopping centers are described. Although many public officials are certainly aware that development is occurring, the substantial scale might be surprising to some. And central city officials dependent on sales and business taxes, as well as other standard revenue sources, need to be fully aware of the rapid changes and ramifications for their cities. This analysis is important because many central city planners and public officials are unaware of the discrepancies in sales, of how communities vary, and more importantly, of the reasons for the continued retail decline and opportunities for the future. Research Outline The literature review is incorporated in three chapters. Chapter 2 provides the status and trends in the discount store and shopping center industry. Chapter 3 highlights the spatial and central city development perspective. The chapter includes investment and portfolio strategies. Then, central city trends and life cycle theory 6 are described. The rationale for government intervention and barriers to redevelopment are identified, including the urban political context for development. Chapter 4 establishes the theoretical foundation for the formulation of the model, including a series of similar research endeavors over the decade on retailing at the metropolitan level. Chapter 5 identifies the proposed model, hypotheses, and methodology. Chapter 6 describes the model's test evaluation and discussion of the regression. And Chapter 7 presents a summary and conclusions including recommendations for future research. CHAPTER 2 LITERATURE REVIEW-PART I DISCOUNT STORE AND SHOPPING CENTER STATUS The literature review is in three chapters and starts with a description of the discount store and shopping center industry in order to appreciate the scale of this important part of the economy. Spatial and financial investment strategy is briefly discussed followed by the positive and negative factors affecting growth of retailing in suburban and central city areas. The economic and political rationale for cities encouraging retail development is highlighted including the important sources of revenue to municipalities such as sales taxes that retail developments frequently provide. Finally, the model building theory will be established including a summation in tabular form of the factors affecting retail spatial location from over fifty journal articles. A glossary of select technical retail terms is incorporated herein, prior to the bibliography. Abbreviated trade journal titles for references as used in the industry are as follows: CSAE= Chain Store Age Executive DM=Discount Merchandiser DSN=Discount Store News LUD=Land Use Digest SCT=Shopping Centers Today SCW=Shopping Center World Historical Evolution of Discount Stores The discount department store industry evolved after Worid War II following suburbanization trends and related household and total income growth. America had pent- up demand for retail goods after the Depression/War era as America's economic fortune grew with post-war prosperity. 8 "But the beginnings of the industry actually began at the turn of the century when 'undersellers,‘ the discounters of their day, sold apparel and soft goods at low markups that undercut conventional department store prices (DSN 21 September 1991)." While many of today's companies evolved from small ventures as far back as the 1920s, most started in the fifties and early sixties using various inexpensive sites in order to substantially reduce facility costs. Conversion of mills in New England was a popular early approach as manufacturers relocated to the south thereby providing inexpensive vacant space near operating mills that were able to supply apparel and domestic goods at low price (DSN 21 September 1991). Variety stores, specialty retailers, conventional department stores and supermarkets were also testing new markets for discounting in this era. Interestingly, four key discount chains all debuted in 1962: 8.8. Kresge with Kmart, Dayton with Target, F.W. Woolworth starting Woolco, and an experienced retailing entrepreneur with experience at J.C. Penney and Ben Franklin by the name of Sam Walton debuting Wal-Mart. The concept of large stores containing both hard lines and apparel, with most sections company-run, was established by that period. In the mid-fifties, discount department stores seemingly posed no threat to the traditional department store chains, but they amazingly surpassed department stores in total sales by 1965 (Bluestone etal. 1981). Key challenges to the discount store industry included Sunday Blue Laws eliminating shopping on Sundays and fair trade laws. "In the 1930s, the Robinson-Patman Act (1936) and the Miller-Tydings Act (1937) were passed. Both of these laws were aimed at protecting small retailers after the Depression to the detriment of chain store operations. The Robinson-Patman Act was enacted because of the discount A&P was getting in the purchase of its products. The Miller-Tydings Act (fair trade) attempted to limit discounting on the part of large retailers by forcing all retailers to sell fair-traded items at the same prices. The fair trade law has now been removed in all states (Berman and Evans 1989)." Manufacturers were obviously supportive of these laws since they were able to limit their 9 supply of merchandise to firms charging the manufacturer's suggested retail price. This excluded discounters who were then forced to carry products with unfamiliar labels and unknown quality (Bluestone et. al. 1981). And the consumer was effectively prohibited from shopping for the best quality merchandise at a market-driven price. Yet, laws do exist that generally prevent suppliers from offering unjustified discounts to large retailers that are unavailable to smaller ones. Bluestone et. al. (1981) state that ”the repeal of fair trade laws may not have had a significant effect on the average price of merchandise, but it had a profound effect on the department store industry as a whole. Repeal precipitated a virtually total restructuring of the retail sector. Abolishing fair trade laws allowed discounters access to national brand-name merchandise that the conventional department store had traditionally monopolized.” While firm consolidations occurred, Kmart's strength in the industry was evident by 1969 as Kmart had over 300 units, more than double the next largest chain. The late sixties and early seventies saw the conventional retailers enter the discount market following Dayton’s success with Target as an upscale discounter including May Company's Venture (no longer owned by May) and Rich's with Richway. In 1976, Kmart added more than one new store every working day (Bluestone et. al. 1981). The recessions of the seventies caused substantial bankruptcy among discount firms, a trend continuing into the eighties with mergers and acquisitions. Such stalwarts as Arlen's, W.T. Grant, Zayre, Alexander's, and Gemco have all disappeared. However, 55 chains founded by 1962 were still in business thirty years later. Reviewing the failure of W. T. Grant and Arlan's, Bluestone et. al. (1981) state that the high failure rate among independents can usually be traced to factors beyond the control of even the best management team including inadequate capital, a lack of scale economies, no access to quantity discounts, and central city location of stores. Undoubtedly, these failings of central city locations are still in the minds of decision-makers today. 10 Technology was instrumental in creating this rapid growth through the use of computers, electronic registers, UPC bar codes, POS scanning and satellite communication systems. Factor substitution in the distribution process has been key to its evolution, especially with changing technology, reduced labor cost through more part-time employment, and economies of scale by managing larger more efficient units. Bankruptcy and consolidation in the early 1980's saw the demise of such discounters as Woolco, FedMart, Gemco, Twin Fair, J.W. Mays, Zody's, Kings, Heck's, and T.G. & Y. By the 1990's, Zayer, Ames, Brendle's, Rose's, Jamesway, and Hills declared Chapter 11 bankruptcy whereas substantial growth by Wal-Mart, Kmart, and Target successfully evolved into truly national operations while other competitors specialized in regional markets (DSN 21 September 1992; 18 April 1994; 21 February 1994; and DM February 1994). More than 50 major discounters liquidated, filed Chapter 11 petitions or were acquired by other firms in the past two years and further casualties are likely (DSN 15 May 1995, 39). Jamesway, Caldor, Bradlees each filed for Chapter 11 in the last six months and Hills Stores is on the block (SCI’ January 1996, l). The once-vibrant 90 unit regional Jamesway filed for Chapter 11 for the second time in two years (DSN 6 November 1995, 1). Consolidations are the seeming natural state of the business with the continuing advent of new types of discounters with low mark up and fast turnover in a service environment attempting to get established and capture market share . Bluestone et. al. (1981) suggest that the discount industry is following the restructuring trends of the manufacturing sector as there is increased concentration in ownership, the growth of centralized financial control, the development of a managerial hierarchy, an effort to reduce the skill requirements in the labor force, the substitution of capital for labor, and the emergence of giant firms that are able to buy out or drive out their competition. Most of the survivors are regional and specialize more than national chains. Furthermore, they tend to be more closely held or private corporations, which have been 1 1 driven by a founding entrepreneur and offspring. Dun & Bradstreet, in partnership with G.A. Wright, Inc. released a study that refutes a 1990 forecast by Management Horizons that predicted half of today's retailers will be out of business. But looking back at DSN's 1990 list of the top 150 chains indicates that 25% are no longer in business, having been shuttered by creditors or gobbled up by competition (DSN 3 October 1994, 14). In addition to the generic ”life-cycle theory," Malcolm P. McNair's ”changing wheel of retailing“ describes the evolution of the retail industry as retail innovators first appear as low-price operators with a low-cost structure and low profit-margin requirements. Over time, they upgrade their products and improve store facilities and locations. As the firm matures, the older institution structures itself to optimize old markets with older facility improvements rather than new formats, becoming complacent while new entrants successfully experiment and innovate (Berman and Evans 1989). The chain store form of organization is a century-old evolving originally due to a modest, homogeneous income consumer base. J.C. Penney, the "five and dimes,“ and early 'A&P" stores were able to handle relatively simple merchandise options and supervision could be centralized. But modern chain operations must have the new capacity for continuous adaptation given market segmentation. Firms have evolved into umbrella operations encouraging the establishment of new experimental chains with new merchandising thrusts. The computer and Whalind management have yielded the opportunity to local marketing even within a large-scale merchandising format. With national expansion slowing down, passed over locations are being reviewed in both central cities and suburbs. And specialty stores and category killers are also moving into suburban highway locations in new power center locations without supermarket or department store anchors. The following describes these shifting trends of development. 12 The Discount Indusg in 1992 The discount store data used in the proposed model are from 1992 and the industry is described for that period for analysis. Additional information through the end of 1995 is presented in the text as appropriate. Whereas mass merchandisers Sears, Penney, and Ward still create 50% of the sales of all full-line discounters, their market share has been declining substantially as Wal-Mart, Kmart , and Target have become powerful competitors in retailing. These "Bi g Three" discounters had 1992 sales of $73.3 billion representing 6.3% of all retail sales, excluding automotive, gasoline, and food establishment business. Their sales represent 65.8% of total full-line discount store industry sales. These three discounters had 42.4% of the total 11,000 discount stores in 1992 compared to 36.6% of the 8,217 stores in 1983, when they were generating only 37.1% of total sales. Kmart has had more than 2,000 stores since 1983 and Wal-Mart operated 1,880 full-line discount stores by the end of 1992. These three companies developed economies of scale and have been better capitalized and more efficient than many of the industry's smaller operations due to the use of new technologies reducing their costs of selling and overhead expenses in order to offer lower prices. New stores are typically located in growing neighborhoods that attract the desired mass custorner--homeowning families with young children. The "Big Three" stores captured two out of three dollars spent at full-line discount stores in 1992. Direct competition among the three leaders was rare until the 19808 with smaller communities in the south to Wal-Mart, with Target in Minnesota and Oklahoma, and Kmart nationally-oriented. Then, Wal-Mart ringed the exurbs and suburbs of the metropolitan South as well as several acquisitions that put them into direct competition with Kmart. And Wal-Mart established a distribution network along the Interstates that set them up for further expansion and market penetration. Target's upscale discount market developed a national strategy extending from the upper-midwest through acquisitions of Ayr-Way's, FedMart, and Gemco in Indiana, Kentucky, California, and the west--then, to the south 13 with the acquisition of Richway stores. Wal-Mart then competed with Target's major, more mature markets (CSAE Special Issue 1993). Now these firms are truly national in scope. Given the concentration of activity in the Big Three and the relatively small number of large retailers, the discount sector is effectively an oligopoly dominating the industry. A high degree of interdependence exists and can even be observed in their agglomeration tendencies at major roadway nodes. And high barriers exist to entry with substantial competition for smaller and regional firms. Firms have implemented corporate strategies through mergers and acquisitions such as Wal-Mart's purchase of many Pace Warehouse Stores from Kmart and Woolco of Canada Store size has increased substantially in order to offer a large quantity of goods and a more creative merchandising presentation. The average new store built in 1992 for the total discount industry rose to 75,000 square feet with over 63,000 square feet of sales area. Major discount department store sizes also increased with most new stores over 100,000 square feet. A discount store is a 'departrnentalized retail establishment utilizing many self service techniques to sell hardgoods, health & beauty aids, apparel and other soft goods, and other general merchandise. It operates at uniquely low margins, has a minimum annual volume of $1 million, and has at least 10,000 square feet of total space (DM June 1993).” At the end of 1992, 9,562 full-line discount stores were in operation with sales of $106.2 billion. See Table 1. The average existing store size was 69,009 square feet with $11.1 million in average sales, for sales of $161 per square foot . Furthermore, the average store includes 10,253 transactions weekly showing the substantial number of consumers who are benefited by such facilities. Note that the average store size for all new discount stores is roughly 9% larger at 75,314 square feet. See Table 2. 14 Table l SaleeAndNnrnberofDiaconntStueaByStoreSize. 1992 — Store Gm. Sq. Ft. N. Store: Sales Sales (‘5) (Billion 3) (‘12) 10,000-24.999 994 10.4 2.91 2.7 mono-49.999 2,192 22.9 13.22 12.5 SEW-74,999 2.691 28.2 27.37 25.8 75.000993” 1.922 20.1 27.82 26.2 100,000+ 1.763 18.4 34.88 32.8 Total 9.562 “5.20 Sm: DM. June 1993 Table 2 DiaconntStoreProfile. 1992 = Category All Discount Stores New Discount Stores No. of units 9,562 360 Total area sq. ft. 69.009 75.314 Sales area sq. ft. 59,244 63,849 Annual sales (3) 11,106,556 9,719,271 Weekly sales transaction (8) 213.587 188.901 Average transaction (3) 20.83 19.10 Weekly I transactions 10.253 9.891 Saledgrom sq. It. (3) 160.94 129.05 Salealaalee area sq. ft. (3) 187.47 152.22 Saves: DM. June 1993 Table 3 indicates that 353 discount store companies existed in 1992 with 177 of these having only single store units. However, a mere 27 firms have more than 50 stores each with 8,150 stores or 85.2% of the total stores and 85.5% of sales. The 3,685 stores over 75,000 square feet represented 38.5% of stores, yet captured 59% of total discount sales. In 1992, 9,831 households per discount store existed in the United States with an average annual household expenditure of $1,129 in discount stores. Certainly, these stores are a major competitive force in the retail environment 15 Table3 SalesAnrlOwnerahipOfDiaconntStoreaByCompanySize, 1992 Co. Sim Sales Sales No. Of Firm Firm No. Of Stores Stores (31111011! 3) (‘5) (‘5) (‘5) SingleStore 1.46 1.4 177 50.1 m 1.9 2-3 1.70 1.6 64 18.1 152 1.6 4-10 4.46 4.2 61 I73 419 4.4 11-25 3.36 3.2 16 4.5 312 3.2 26-49 4.38 4.1 8 2.3 352 3.7 50+ 9083 85.5 27 7.7 8.150 85.2 Total 15.20 3Q 9.562 Sartre-e: DM. June 1993 Table 4 highlights the key categories of sales within discount stores, illustrating the importance of apparel goods in the discount format Although these department shares are for all discount stores, it should be noted that the f ull-line discount department store has been increasing its sales share in the apparel category as upgraded brands and extensive marketing have been pursued to increase higher margin apparel sales. Table 4 Salaa Share By Major Departments. 1992 pep-mu Sales Total (Billionfl) (‘5) Apparel 3294 31.02 Hardlinea 27.65 25.04 leisure Gooda 18.08 17.02 Variety 9.93 935 l-IomeFaahiona 6.70 631 Drup&Ccarneticl 9.05 1152 Minoan-eon 1.35 1.74 Total 106.20 Source: DM. June 1993 Table A.1 provides a listing of the top discount store chains with greater than $50 million in sales. These 54 firms represent a major portion of sales and are shown to 16 highlight the diversity of sizes and regional locations of coverage. Many smaller regional firms exist in the discount store industry although consolidation is increasing. The data also show the tremendous strength of the ”Big Three” in influencing the discount industry, with 4,679 of the total 9,015 in 1992. These stores represent 51.9% of the total, but capture $69.9 billion or 65.8% of total sales. W Professor Douglas 1. Ti gert, Charles Clarke Reynolds professor of retail marketing at Babson College, completed a survey sponsored by Chain Store Age Executive of the Atlanta, Indianapolis, and Dallas/Fort Worth metropolitan markets regarding market information for the major discounters (CSAE Special Issue 1993). Nearly 90% of the population shopped in the major f ull-line discount stores thereby indicating that discount stores have evolved virtually into the department stores of this generation. Those who did not qualify for the survey (anyone who did not shop any of these three stores at least once over a two-month period) were bipolar, either under age 25 or over 65 years of age; and the non-qualified were more represented in those with incomes below $20,000 or above $80,000 per year. Tigert indicates that lower-income, young households cannot afford to shop that often at any kind of store, while older and/or higher income people may well shop elsewhere or less often. It is important to note, however, that shoppers surveys are indeed affected by the nature and location of the competitive supply. The implication is that these people do not shop in the major discount stores; one could suggest that stores may not be located in these areas thereby reflecting less consumption by these household types. This constraint of market surveys may be a key problem for central cities if the supply of stores is not available in the central city to retain shoppers. Since central cities tend to have an older population, for example, are discount store location analysts ignoring potential opportunities because their shoppers surveys in the suburbs (where population is younger) would show that they cannot attract the older shopper? Similar perspectives could be an 17 issue for minorities and lower income households since they are more represented in the city, yet underrepresented in market surveys. The general market survey may be correct, but perhaps central city analysts should determine analogs from central city environments to use in marketing efforts for new development. Although markets vary somewhat even within the discount store category, Kmart's primary shoppers were more likely to be unemployed or working part-time, older, and to be in lower income households. Target shoppers tend to be more from executive/professional ranks, and in the 25-44 age group. Wal-Mart shoppers are willing to travel the furthest with roughly one-third traveling more than five miles and only 15% of Kmart shoppers and 20% of Target shoppers traveling that distance. Wal-Mart tends to have higher sales per customer, due partly to much more frequent shopping with 37% visiting Wal-Marts weekly in Dallas and Indianapolis. One can see the different approaches in competitive strategies among the various discounters trying to become noted for cost leadership, differentiation of merchandise and presentation, and focus on market segments. As stated previously, all these different trends are a function of distance, location, and competition; also, these newer markets may include food sales from supercenter units maln'ng the stores more attractive in newer formats. Tigert hypothesizes that "consumers choose the store where they shop most often on the basis of which store is perceived to be the best (even if only marginally, but noticeably better) on the largest number of store dimensions that are important to that consumer. For any consumer segment, there are only two or three critical store-choice characteristics that drive the store-choice process." Tigert concludes that discount store choice is similar to that for supermarkets-location, price, and assortment. He indicates that an effectively competing store must win on at least two of these categories to generate the highest sales per square foot and high gross margin dollars per square foot. Tigert indicates that satisfying all three factors may be counter-productive as he indicates that convenience may not allow for higher sales or gross margins per square foot. 18 A retailer should be spread out enough so that low-price appeal can draw from a larger radius. He also indicates that shoppers have a two-stage selection process focusing on convenience first and then looking at price and assortment. As an indication of national strength, Wal-Mart had sales of $287 per square foot compared to Target's $209 and Kmart's $140 in 1993. Wal-Mart also leads in operating earnings per square foot at $21.53 compared to $11.21 and $7.57 for Target and Kmart respectively. Table 5 reflects the year-end data for the ”Big Three” in 1993 and the superior performance of Wal-Mart. Note that Wal-Mart is outperforming its competitors extensively although Target is performing relatively well. Table 5 Big3 Discount Department StoreTrenda.1993 Wal-Mart Target Kmart Ave. sales/sq. it. (3) 287 209 140 Comparable more sales-last year (1» change) 83 5.5 3.0 3vyr. annualized store unit growth rate (‘5) 8.7 82 1.0 3-yr. annualized sales growl rate (‘5) 23.8 11.4 2.1 Avestoreaalea(millions$) 24.9 21.9 113 Gram mal’lil (‘5) 235 24.3 24.6 Expense ratio (‘5) 16 18.9 19.2 Operating income per ave. sq. ft. (S) 21.53 11.21 7.57 Sources: Montgomery Securities; Discount Store News. 18 April 1994. 9. Coverage of stores and market dominance are not necessarily the best financial strategies as Kmart has a substantially larger number of stores in these markets. In Indianapolis, for example, only 15% of Kmart's primary shoppers drive more than five miles compared to 38% of Wal-Mart's. And more than 40% of Kmart shoppers travel one mile or less compared to just 14% of Wal-Mart's. Kmart stores may be in the inner ring suburbs or the central city compared to Wal-Mart in the new growth areas in the outer ring. 19 They may need to either cut back and create less number of stores (but supercenters) or build further out to capture new growth markets. From a cross sectional perspective, Kmart's existing store strategy may not be economically efficient for today; yet, historically, prior location decisions and scale of development may have been wise. Perhaps Kmart has older stores that do need replacement due to new market penetration or declining trade areas, yet they may have had a successful twenty-year history of generating income and profit from these stores. As anchor vacancies increase, "aggressively expanding hi gh-volume discounters seeking new markets or larger prototypes in existing markets, aided by low interest rates, cheap land, and more flexible local governments that are desperate for new tax revenue, are finding it more expedient and economical to construct new buildings rather than move into vacant, imperfect 'bi g boxes.‘ In addition, existing 60,0001» square foot tired-looking Kmart stores and their counterparts in other categories must compete with newly expanding 116,000 square foot Wal-Marts and their counterparts, or fail (SCW May 1992, 140)." The next section discusses the state of the industry and store development plans for the major discount stores. State Of The lndusg And The Big Three Distinctions among retail sectors have been altered as retailers expand their offerings to meet consumer desires for convenience and generate more traffic. ”Scrambled merchandising" whereby various store types have diversified their offerings by adding more merchandise lines is changing consumer shopping patterns toward destination orientation for one-stop shopping. The supercenter format for the major discount stores (along the approach of Meijers in Michigan) will further blur the discount store format as these stores become directly competitive with the supermarket industry (CSAE August 1993). Full-line discount stores have $150 billion in sales of a total $318 billion discount market, about 48% (DSN 3 July 1995, 1). 20 Discount stores have evolved from a low-price, low-quality merchandising format oriented f or less-af fluent consumers to a broader market appeal even with upper income shoppers. Discounters have enhanced their apparel offerings and have upgraded the presentation format within their stores such that they are now penetrating department store and national mass merchandise sales while retaining their low price strategy. Table 6 highlights the degree of shopping activity by consumers for discount department stores and category dominant stores such as Office Depot, Supermarket of Shoes, and The Sports Authority. Table 6 Monthly Visitat'nnand Spendingby StoreType W Wm 1223. 12% 1223. 1% Vii! 3.2 4.2 0.8 1.2 Average spending 3% $99 $32 $48 Source: lflmGdlerganizationStudy Given the slower economic climate, consumer behavior patterns have shifted towards convenience, quality, and price thereby being value-oriented. Consumers are also time- conscious with destination stores incorporating large selections within particular product categories. “Value retailing is clearly the theme of the day, and probably the decade, and those retailers that know how to combine low prices with hi gh-quality merchandise, convenience and reasonable service appear to be the ones that will come out of the current recession in the strongest position. New regional malls are not being built and are being supplanted by power centers and factory outlets. The conventional image of a regional mall is already out of date, altered by the addition of hotels, office buildings, libraries, city halls and/or entertainment complexes and by design changes that have integrated some centers 21 much more closely with their surroundings than had been the case in the past (SCW December 1992, 18)." Begun in the mid-e1 ghties, power centers are the new wave of shopping center development. A power center is defined by the International Council of Shopping Centers (ICSC) as a shopping complex of 200,000 to 400,000 square feet or more, at least 75% of which is leased to leading deep discounters, or category killers like Circuit City or Kids 'R' Us. These centers usually have three to five promotional anchors selling both hard and soft goods. From 1988 to 1991, the ICSC estimates that powers centers increased in number by 22% with more than 2,500 by 1993. Many shoppers do little cross shopping but are attracted to the site because of convenience, value, efficiency, and reliability. Over 85% of shoppers actually buy something at a power center compared to 50% in a mall (SCT November 1989; Women's Wear Daily 13 November 1992). Power center location analysts look to areas with increasing household formation, population distribution, and job growth. From a rent structure standpoint, an anchor will pay 60-70% more than in a smaller center due to the increased volume potential (Stores March 1989). Locations tend to be on or near major highways or their arteries. Also, many older centers are redeveloping into this concept. Consumer buying habits have changed with less loyalty to department stores than in the past. Consumers shop for the best values. Regional malls still offer atmosphere and entertainment opportunities, food courts, etc.-more of a shopping experience. And income stability can be stronger in a regional mall than in a power center if an anchor pulls out thereby affecting the value of whole center (SCW June 1993, 66). However, major declines in shopping frequency from 1988 to 1992 can be seen in regional mails from 52% to 41% and urban shopping districts from 38% to 26%. These represent primary shoppers who shop once a month or more for non-food/drug items. Only neighborhood convenience centers have held constant at 50%. Discount department stores have also held steady during the period at 60% whereas a major decline has occurred 22 for traditional department stores such as Dillard's and Hudson's from 35% to 20%, J.C. Penney from 21% to 16%, and Sears with substantial decline from 20% to 12%. Since the mid-19808, the biggest winner across all merchandise categories has been the discount department store (CSAE August 1993). A key reason for the growth of the discount department store as a retailing mainstay for the middle-income market has been the significant upgrading of apparel goods from a fashion, quality, and availability perspective. These improvements have cut into sales at the traditional department store, national mass merchandise, and apparel store levels. Since 1988, preferences for women's casual clothing increased from 23% to 32% in discount department stores and from 29% to 44% for children's clothing. Perhaps the vitality of some of the leaders in the discount industry can be shown by a 1992 study in Table 7 of consumers who shopped in various stores the previous year. Table? PercentageofStoresShoppedinbyU.S.Consumersin1991 Store Percentage of Shoppers Kmart 77 Sears 67 Wal-Mart 51 Target 42 Toys R Us 42 Montgomery Ward 29 Sam's Club (Wei-Mart Owned) 24 Mervyn's 20 Amos 16 Pace (acquired in 1993 by Sam's) 12 Solace: DSN. 16 March 1992. The ”Bi g Three” are leading the industry in new formats and growth. However, Kmart has recently announwd a major revision to its development plans that are described 23 in detail as follows. Kmart's revised development program may have significant economic implications on communities, given the move away from their older, obsolete units. W Kmart had uninterrupted success in the sixties and seventies with its successful homogeneous development formula. According to Arthur Markowi tz, Senior Editor of Discount Store News, they fine-tuned rather than fully studied their formats through the years. He stated that management became inbred, insulated, and unwilling to either study its competitors or consider experimenting. Their initial desire to remodel 2,000 old stores was altered by consumers passing them by to find newer centers with state of the art facilities and merchandising. Markowitz believes Kmart made a strategic mistake by trying to diversify its business into other retail store formats and lines (e. g. Pace, Builders Square, BizMart, and others) instead of dealing with the existing problem of functionally economic units. Now, it has decided to directly face its problems, close older, inefficient and unprofitable stores, remodel viable units, and invest heavily in supercenters (DSN 21 February 1994, 12). Kmart has adjusted its store renewal program begun in 1990 to include more closings of smaller units, which are struggling against mounting competition, as well as to accelerate the rollout of the more successful 110,000 sq. ft. discount store prototype and the even larger Super Kmart Center food/general merchandise format. In 1990, Kmart embarked on a $3.5 billion refurbishment program for 1/2 of their 2,400 units with flat comparable store sales and declining profits. New 1994 plans included 800 relocations up from 300 in the 1990 plan, 700 expansions up from 620, and 670 ref urbishments down from 1,250. Also, they will do a complete renovation of their Canadian store operations (Wal-Mart recently acquired Woolco in Canada), upgrading of Builders Square 11 format, and other costly changes to their original $3.5 billion store renewal program. Kmart's attempt to sell off 25% stakes in Builder's Square 177 units, OfficeMax 330 stores, Waldenbooks 1,247 stores, The Sports Authority 81 units, and Borders 82 stores 24 was voted down by shareholders (DSN 21 February 1994, 51; DSN 20 June 1994, 1).These changes are due to major financial decline and weak performance as illustrated previously in Table 6 (DSN 21 March 1994, l). The dual effects of a 32.3% decline in operating income for Kmart discount stores and the contraction has left it strapped for cash to complete the renovations with more than 1,000 remaining to be remodeled, relocated, or replaced with supercenters (DSN 16 May 1994, 3). Kmart sold 91 of 113 Pace warehouse club stores in 1993 to Wal-Mart for $300 million after an $87 million operating loss. Thus, Kmart has sold off its specialty stores to concentrate fully on the ailing discount store business. The main issue is the conclusion of the renewal program with half the chain dead or dying, according to David Poneman of Sanford Bernstein & Co (DSN 5 September 1994, 1). In September 1994, Kmart announced it would be closing 110 discount stores and fire 10% of management over the next two years. They announced the closing of the stores during January/February 1995, creating the loss of 6,000 store employees. Chairman Antonini insisted that all stores fully meet return on investment requirements (DSN 19 September 1994, 1). Table 8 shows the 1994 change in strategy by Kmart to deal with the problem of older inefficient units. Updated Kmart stores outperform unrefurbished units by 17% in sales, 12% in customers, 6% spent per visit, and 35% higher profits. Kmart then operated a total portfolio of 4,341 stores (4,182 in the domestic 50 states and 159 in Canada, US Virgin Islands, Puerto Rico, the Czech Republic, and Slovakia, 2,430 of which are Kmart discount stores (DM February 1994, 18). Studies do show more shopping and more dollars at renewed units, although Wal-Mart is still tallying aggressive gains on Kmart shoppers with less sales and operating profits per square foot than either Wal-Mart or Target. These revised plans may have significant impacts on central cities and inner ring suburbs with smaller, functionally obsolete facilities, typically in the 40,000 to 65,000 Table 8 Kurt's Revised Development Program . 1990-1996 Type 1990 Plan 1994 Current Status 1996 New Plan Relocation 300 300 8a) Expanl'ons 61) 430 700 Refurbishmeuts 1.250 450 6‘70 Source:1(martCorp.Jan.1994 square foot range. New units are around 110,000 square feet. For example, Kmart recently added two new units to five existing Super Ks in the Chicago area but they closed seven, smaller older discount stores due to the openings (DSN 21 March 1994, 4). The positive impact of a store relocation can be substantial (albeit with more costs) as illustrated by their new Super K in Medina OH. that replaced an existing obsolete store thereby increasing sales from $12 million to $80 million in 1993 (DM December 1993, 16). Decisions to close stores also have an impact on market share of other discount firms with 500 store closing/relocations to leave 69 Wal-Marts and 10 Targets without competition (DSN 3 October 1994, 1). Kmart is attempting to build 500 Super Ks by 2000. Kmart executives indicate they intend to have 500 supercenters by the year 2000, compared to 80 today in the 120- 190.000 foot range (DSN 4 September 1995, 59). Super K generates sales of $350 per square foot on the general merchandise side of the store, double the typical unit, with total sales of $50 million estimated from the 75 units to be in operation in 1994. Kmart has yet to compete with another super center. Some analysts believe that only one supercenter concept per market can be successful with the first in the market having the best success-- similar to the warehouse club penetration. Kmart plans to open 55 Super Ks with plans for a total of 70 by '94, 125 by '95, and 190 by '96 (DSN 21 February 1994, 51). As an indicator of the dynamism of the discount decision-making process, by October 1994, the timetable for supercenter expansion was accelerated, with 75 more in the 1994 fiscal year and at least 100 in 1995 for 243 projected by the end of 1995, with plans to add 26 100-115 per year thereafter. Expansion of relatively new 110,000 foot units will be key and 84 of the 100 1995 supercenters were proposed as conversions or relocation of older 80,000 foot units (DSN 3 October 1994, 1). By Spring 1995, however, progress was deemed inadequate when long-term Chairman Antonini was replaced by f orrner Target and Grand Union Supermarket executive, Floyd Hall. Even potential bankruptcy was discussed through January 1996, when financial restructuring seemed to deflect immediate concern of financial solvency. Kmart indicated during 1995 that it would close 196 discount stores this year exiting 71 markets and reducing its presence in 72 others. During 1995, store expansion was slowed to 43 new discount stores and 22 new Super Kmart Centers. They are also considering smaller box prototype for Super K that would be appropriate for smaller markets (DSN 1 May 1995, 1). Kmart plans to liquidate leases in 205 shuttered stores in 35 states through an auction with terms ranging from two to twenty-three years. Stores range from 40,000 to 135,000 feet with 73 in the midwest, 49 southwest, 47 southeast, 27 west, and nine in the northeast. Texas, California, Florida, Illinois, and Michigan each have 10 or more in the auction (DSN 3 July 1995, 2). Kmart operated 18 power centers in 1993 with five more planned for 1994. These large-scale centers are anchored by ”category-killers" and have a major impact with over a five-mile drawing power. Although their concept originally was to have primarily Kmart owned stores, they are still interested in the concept even if they are not all their stores (DSN 6 December 1993, 96). For example, Kmart and its subsidiaries stores operated over 400,000 square feet of the 500,000 square foot power center in the city of St. Louis, assisted by tax increment financing. Developed by Midland Properties in 1992, this $52 million shopping center included Kmart, Pace, and Builders Square, Biz Mart, and Supermarket of Shoes. Recognizing that market segmentation is vital to future retailing, Kmart is trying become more competitive for minority shoppers and has hired Leo J. Burnett to develop ad 27 campaigns aimed specifically at African-American and Hispanic audiences. Senior Vice- President of Marketing, Mike Wellman, indicates that "both Hispanics and African- Americans are more loyal Kmart shoppers than any other group. They tend to be somewhat younger and have children; that's our customer base.” Radio and television ads have been produced for stations serving those markets and for national use, and stores with larger minority shopping bases have been nricrornarketed slightly to appeal better to those shoppers. Kmart has become a leader in cosmetics for African-Americans and its apparel department shows distinct shifts, such as increased selections of leather jackets, in urban settings. An IRMA Conference on Diversity in the Marketplace indicated that minority consumer groups are growing rapidly. African Americans have gross national income of $280 billion according to the 1990 census with half of the market aged between 25 and 34 and 56% women. Furthermore, blacks are relatively easy to reach since 56% live in the 50 largest metro centers (DSN 20 June 1994, 12). Kmart still is perceived as being more expensive than Wal-Mart by consumers. They are being somewhat undercut by Wal-Mart on value position at the low end and not adequately competitive with the higher end of the discount spectrum. Target, ShopKo, Fred Meyer, and Venture have moved upscale to appeal to the former department store shopper. Yet about half the US. population walks through a Kmart every month and they are looking for more fashion in apparel. Furthermore, Wal-Mart managers apparently have more power over their own stores catering to individual markets with more of a bottom up strategy devised by Sam Walton who tried to empower everyone versus top-down administration (DSN 21 February 1994, 52). Wall Street questions the restructuring effort and suggests maybe it is too little, too late. It is as if "they suddenly woke up and found that a third of their discount stores were obsolete and Wal-Mart was breathing down their necks. They knew that five years ago," said Linda Morris, retail analyst for PNC Bank, Philadelphia. Wal-Mart and Target operate better, fresher stores in better locations, says Marvin Behm for Duff & Phelps, which has 28 Kmart on credit watch. He says that Kmart's competitors are also technologically superior identifying what will sell, buying it, and getting it distributed to stores and moving it out if it fails to sell. Wal-Mart tries one markdown then eliminates it while Kmart tries several markdowns. Kmart is still operating a number of 60,000 foot stores, dragging its average down to 90,000 feet. Sales per square foot have continually been declining from $189 in 1990 to $179 in 1993 (DSN 21 February 1994, 56). Although Kmart staved off bankruptcy in early 1996, they were placed on credit watch during a critical period of restructuring. They intend to close 70 stores and open 30 new stores and 20 Super Kmart centers during 1996 (DSN 5 February 1996, 3). Wal-Mart Stores Approximately 70 million customers shop weekly at Wal-Mart (DSN 20 November 1995, 4). In 1982, Sears controlled 26% of the sales for the country's top ten retailers, with a seven point lead over number two, Kmart. Ten years hence, Sear's position in the market slipped to third in total sales and 13.5% of the nation's sales. According to Citicorp, Wal-Mart held a 30.7% share of sales for the top ten retailing companies in the U.S. in 1993, up from 4.8% in 1983 (DSN 5 December 1994, 58). Wal-Mart has increased from 1.6% of total GAF sales (33.5% in the top ten retailers in 1983) to 12.0% in 1993 (39.0% in the top ten retailers). Wal-Mart represents 14% of GAF sales in the United States, more than any other company in history (DSN 19 June 1995, 90). Wal-Mart recorded a 17% growth in earnings to $2.33 billion as total company sales jumped 21.4% to $67.9 billion in 1993. Wal-Mart Discount Store sales are at $49.7 billion (growth of 22.1%) with McLane Stores at $3.5 billion and Sam's Club at $14.8 billion. In 1994, Wal-Mart planned to open at least 125 discount stores and expand or relocate 255- 275 Wal-Marts including 68 Wal-Mart Supercenters in the U.S. and add 20-25 Sam's Clubs. Internationally, Wal-Mart is in a very aggressive mode acquiring 120 Woolcos in Canada. And its venture with Cifra includes plans to open 38-40 stores including 1015 Sam's and supercenters in Mexico. In 1993, the firm opened 142 discount stores and 29 supercenters, expanded or relocated 130 Wal-Marts, and added 163 Sam's Clubs including the 99 Pace Membership Warehouses purchased from Kmart (DSN 7 February 1994, 1; 7 March 1994, 1 ). Not only are Wal-Mart's stores newer, they are much larger too averaging about 125,(X)0 feet, up from the 1108 that dominated just five years ago. Most stores have been designed with breakaway walls and extra land. Where a store is an obsolete 50-60,000 sq. ft. model from the '703 and early '80s, newer stores are built nearby and convert the unit to a Bud's closeout operation, rent it to another non-competing retailers, or raze it and sell the property. Wal-Mart's architectural firm has developed a computerized system that adapts the basic box to local conditions and regulations, speeding up the approval process and getting construction underway much faster than at competing chains (DSN 5 December 1994, 71). Gene Wright, retail industry director for Andersen Consulting, claims that "the key buzzword is planned organizational obsolescence: destroy yourself before the competition does (DSN 15 May 1995, 43)." To give a sense of scale of the warehouse distribution nwds for a major discount 1' 11m, Wal-Mart has 27 distribution centers. Even these have been assisted occasionally through public-private partnerships. They are attempting to work with the Industrial Development Authority in Sharon Springs NY. for a $30 million, one-million square foot facility on a 217-acre site to be bought by the IDA and leased to Wal-Mart (DSN 17 January 1994, 1). Redeveloping closed facilities is a major endeavor by either the discount firm or the shopping center developer with vacant space. Many of Wal-Mart's closed stores become Bud's Warehouse Outlets to handle surplus goods and fresh goods close-outs. Wal-Mart operates 77 Bud's and expected 30 more in '94. Founded in 1990, Bud's generated $200 million in 1993. They are also trying different formats like a new farm store concept in Kirksville MO. (DSN 6 December 1993, 96). 30 As of Jan. 31, 1993 they operated 1,954 Wal-Mart stores, 68 Supercenters, and 419 Sam's employing 520,000 associates (DM March 1994, 24). Wal-Mart's David Glass recently spoke with top investment analysts and expressed admiration for convenient one- stop shopping like Dollar General and Family Dollar, suggesting a potential market opportunity. Wal-Mart is within 5(1) units of total penetration of the U.S. retail market with diminishing returns afterwards unless new types such as convenience goods develop (DSN 6 November 1995, 3). Sam's Club's growth from 256 units in 1992 to 419 last year and at least 439 this year will be a major element in the overall jump in sales (DSN 20 June 1994, 75). By the end of the decade, Wal-Mart stores will cease growing on a net store basis due to the advancement of supercenters and the continued conversion of many discount stores into the supercenter format. Wal-Mart's sales are $300 per sq. ft. GLA, double that of Kmart (DSN 5 December 1994, 67). Perlmps, all good things must really come to an endueven Wal-Mart's spectacular performance over 99 consecutive quarters of increasing year-year earnings toward $100 billion in annual sales. Year end 1995 results showed a decline in the 100th quarter but expects $95 billion in sales with $2.7 billion in earnings. Domestic plans for 1996 include converting 95 Wal-Mart stores into supercenters, building 15 new supercenters, opening 75 Wal-Mart discount stores, and opening 12 Sam's Clubs (DSN 5 February 1996, 1). mm Target, a division of Dayton-Hudson, planned to open 65 stores in 1995 on top of 57 new stores in '94, mainly in the 115,000 square feet prototype (DSN 19 June 1995, 21). Target will open 80 stores in 1996 with a total of673 stores in 33 states at the end of 1995 (DSN 1 January 1996, 51). Total sales for Target went up 16.2% to $13.6 billion with 623 stores for sales of $222 per square foot with an average store size of 103,700 square feet. Over the next two years, Target will open around 50 new stores in DC. and Baltimore; Target also plans for 31 new stores in Philadelphia to enhance its Baltimore/DC. strategy. Target likes to open stores close to upscale malls and likes to be 3 1 near Wal-Mart for competition. Target's penetration of the Northeast is expected to be a fresh approach to business, yet its arrival in the overstored and under-profitable territory should require some consolidation of Hills, Roses, Clover, Caldor with additional harm to Kmart (DSN 1 January 1996, 15). Mr. Ken Woodrow, Vice-Chairman of Target Stores, states that Target's new vision is "Total Quality." They have altered their market pricing strategy to every day low pricing (versus advertised sales markdown techniques) and they have a mission to become an active participant of the community. Their plans emphasize major improvements of technology and improved management systems. Similar to Wal-Mart, they intend to share more data with suppliers. Target prefers to own their own property and they have very few they would like to relocate believing that the underlying land has future value as well. Target has recently exhibited its new prototype for its full-line discount stores with the Greatland stores at 125-130,000 square feet. The conventional Target is 115,000 compared to 125,000 for Greatland, which has two entrances. Their expansion strategy has been to pursue successful acquisitions to enter markets. After their recent movement into Chicago and Cleveland, they are now expanding in the Northeast. Target has a corporate objective to find locations for a desired 10% annual growth in space. Los Angeles has 60 stores and 113 in Califomia as its largest market. Given the broad ethnicity of its markets, Target is pursuing substantial training and college recruiting. Market segmentation is vital in their strategy as they are trying to focus on micro-marketing working closely with Hispanic, Asian and African-American communities. Woodrow indicates that they do not see themselves as the largest discount chain in the nation but as the premier discounter-the best-the preferred alternative (DM March 1994, 38). Target is trying to fight and win a two-front merchandising strategy including penetrating the Northeast where Caldor and Bradlees have carved out the upscale discount market. And they have simultaneously lowered their prices to compete directly with Wal- Mart on common items. Target is committed to micro-marketing each store. 32 They are late entrants in the supercenter format and have developed the new Target Supercenter to look like their new Greatland discount store format, but with a f ull-line supermarket. Target opened its first Super Target in March, 1994 in Omaha at 195,000 square feet with a smaller prototype in Lawrence, KS. at 160,000 square feet with around 45,000 square feet in groceries and a bank branch, pharmacy, photo studio and restaurant. Champaign, and Springfield, IL. are also proposed sites. In Iawrence, Target purchased development rights to 24.5 acres in dozens of residential parcels (DSN 21 February 1994, 2). Target also plans 10-20 supercenters over the next two years although this could double if successful (DSN 5 June 1995, 53). Target also looked for a site within the Chicago City Limits. To show the kind of bias that exists in some of the discount store community towards central cities, a recent article by Discount Store News stated that Target recently " opened a store within the Dallas city limits indicating that land scarcity and urban problems, including crime, fail to faze the chain (DSN 18 April 1994, 25) ." Given national store expansion by the "Big Three," regional discounters have been cutting costs, downsizing staff, and upgrading stores and technology to become more efficient competitors. Competition and market saturation are forcing discounters to define their position in the marketplace. ”Upgraded store environments; focused, in-stock merchandise assortrnents; on-trend fashions; attention to customer service; and a consistent value offer have become competitive requirements." Urban Market En Trends Given the constraints of land assembly (cost and availability) in large central city and inner ring suburban areas, discounters are looking at multi-level formats primarily in abandoned department stores. Interestingly, the discount store retailing cycle comes full circle with supercenters and multi-level stores that were seen in the formative years of discount retailing. Korvettes originally spotted units adjacent to supermarkets (e. g. a forerunner of the supercenter). 33 Also they had many multi-level units in urban areas emulating Alexander's and Masters. Now recycled department stores are being reused by Kmart, Bradlees, Caldor, Target, and Fedco (DSN 18 April 1994, 12). Fedco opened its first multi-level store during November 1993 in a former May Department Store in Buena Park (LA) , and is negotiating for a second unit in the San Bernadino-Riverside area. Conversion took nine months with a $100 million sales goal in five years. Of 10 Fedco Stores, a members-only chain, only three have reached $100 million. The store is gaining 20,000 members per year with 100,000 already at $5 lifetime fee. A co-operative, 50% of its members are govemment employees. This adaptive reuse has 35,000 square feet of foodstore space compared to 10,000 feet at a standard Fedco. Store design was a major concern and four elevator cars appear to be adequate. The store is the first new store for the chain in seven years and is being leased (DSN 4 April 1994). In March 1994, Target opened its first two-story unit in a former Robinson's Department Store at 163,000 square feet. A normal Target P1 Store requires 11 acres while this store is on only 4 10 acres in Pasadena. A three-level deck for 548 cars is attached with a pedestrian bridge linking it to the store's second story. A year-round garden center is on top of the parking garage. Four elevator cars holding 10 carts each exist in the core. The top level of the 3-level department store is for storage and administrative operations. Target will be evaluating whether operating costs are higher on a three-story building that has inventory stored on top, four entrances instead of one (two at Greatlands) and three banks of checkouts instead of one (DSN 4 April 1994). Urban, inner city locations are now being sought by discounters with the disappearance of Zayres and others through the years. Woolworth, Goldblatt's, and Family Dollar have had successful city operations, although these are not full-line discount department stores. Frances Trachter, Vice-President, Public Affairs for Family Dollar says their urban stores are their most profitable and represent their long term strategy, even though they are closing stores in smaller markets. Goldblatt's has 14 units in Chicago, but 34 nine security personnel are required at the chain's older, multilevel units due to the prevalence of gangs and crime. Family Dollar is creating a minimum of eight basic assortment models to address the tremendous ethnic diversity that exists throughout the 2,070 unit chain. Their computer models try to detemrine product shopping patterns by ethnicity. Kmart is looking more at downtown expansion and has opened a store in Queens and East Oakland, CA. In New York City's Herald Square area, Kmart, Sports Authority, and Caldor are launching a discount store conidor in the famous hi gh-brow shopping district. DSN indicates that discounters are arriving in New York City to fill the last great retail frontier in the U.S.--inner city locations. Kmart is looking for another four sites in the area. They opened its first store in The Bronx in an area known as Co—Op City and operates two stores in Staten Island and one in Queens. Retail observers say such stores are risky since commuters tend to run for their trains when heading home and do not usually stop to shop; distribution logistics; and shrinkage (DSN 2 January 1995, 3). The New York metro market is being evaluated given the sheer density of population and income. Wal-Mart recently pulled out of New York due to prior market entry and penetration by Kmart and Price Club as well as high costs of construction and lack of knowledge of market potential. Bradlees has a store in Yonkers with at least two new ones proposed on Long Island and one in Brooklyn. Now, Bradlees is looking at Manhattan in a former JW May building, becoming Manhattan's first large discount store since Alexander's closed a year ago. Kmart's first Long Island Store did $52 million versus an estimated $8 million and $42 million in Levittown, built across the street from an existing Caldor's. Wal-Mart is looking at developing 12 stores on Long Island. Target has decided to stay out due to the strength of Bradlees and Caldor's. Bradlees has a new 83,000 square foot unit in Totowa. Construction costs are substantially higher at $38 per square foot versus a $22 average elsewhere, but the potential return on investment appears to be higher. 35 Kmart has confirmed it may site a three-level 150,000 sq. ft. store at 770 Broadway at Astor Place in Manhattan, formerly a building that housed John Wanamaker department store fifty years ago. The site is in the midst of the NYU campus area, just six blocks away from Union Square with big box retailers. Near the flagship Macy's store, Kmart is currently renovating 145,000 square feet for its first NYC store to open in 1996 at Penn Station at 34th Street (DSN 4 September 1995, 2). I"The city is interested in maximizing opportunities and retailers are realizing there is potential here they have not tapped," says Angela Brown, vice president for development at the city's Economic Development Corporation that provides assistance including tax abatements. She figures that the outer boroughs of Queens, Brooklyn and the Bronx could each easily accommodate six major discount department stores-a growth projection limited only by available sites. In terms of demographics, she adds, becoming over-stored is ”almost impossible.” Bradlees Brooklyn store is EDC assisted. Also, the EDC has assisted the development of a major mall on Staten Island to be anchored by Price/Costco and a plan to build a major retail center with up to 240,000 feet of store space in Jamaica, Queens. Manhattan has a limited number of discounters with no shortage of city locations for existing space including some prior department store sites, according to retail analysts. Bradlees is interested but is studying its Yonkers store as an indicator of the potential market. Korvettes and Alexander's failed in the past to capture an adequate market and discounters are hesitant to invest where someone previously failed (DSN 6 December 1993, A30). Bradlees urban multi-level unit opened in Yonkers at 144,000 square feet on three floors compared to 83,000 feet for their typical unit. Both Caldor and Bradlees are trying to target their merchandising at upscale discounting although no one has attempted to replace central city lower income specialists like Zayres and Alexander's. They are looking at the Philadelphia through Boston conidor with emphasis in New York. Both store chains have battled for a number of former Alexander's sites. A major city-sponsored 36 commercial development called the Atlantic Center Development anchored by Bradlees including housing has been proposed for a 1994/95 opening (DSN 6 December 1993, 1 10). Manhattan will soon be home to Bradlee's largest store at 143,000 sq. f t.--a seven level former May's Department Store on 14th Street/Union Square to be accompanied by a 43,000 foot, 3 level Toys 'R' Us. Several other reuse projects are being discussed in Brooklyn, Queens, and other New York City markets (DSN 5 September 1994, 4). Bradlee's 136 stores has overall sales per net selling sq.ft. ranging from $200—$230 with Manhattan expected to be in the $350-$375 range. Shrinkage is estimated to be double their norm at 4% (DSN 21 November 1994, 1). Bradlees and Caldor started competing in November 1992 for six former Alexander's stores. Caldor won the competition and reopened three in the fall, 1993. Kmart converted a former Bloomingdale's in Fresh Meadows, Queens. Clover is trying to open a discount store on lower and ground levels of the five-story Stems Department Store closed in 1992-- without parking, although buses and subways are at the bottom of the store (DSN 6 December 1993, 110). Caldor's urban focus has concentrated on Flushing Queens and The Bronx including a store in Staten Island plus three others opening in The Bronx, Brooklyn, and Queens with Bradlees as the main competitor. Two prior Alexander units bought by Caldor in 1992 are multi-Ievel with four floors at Flushing and five at the Bronx. Also, they acquired a 2—floor Macy's site in Flatbush, Brooklyn that was slated to open in 1994. These stores at 165.0(1) and 184,000 square feet are larger than their 113,000 foot average. Sales are anticipated in the $200 range; merchandising and adjacencies had to be completely reconfigured for the units. Caldor has been more willing than most to enter dense urban areas and they are looking at suburban Washington, DC. for two units. Caldor disclosed plans to open its first Manhattan store near the year-old Bradlees unit at Union Square anchoring a $100 million project with a 150,000 foot store consuming abouth the retail 37 space with The Sports Authority in a 20-story building including 200 rental apartments- this announcement comes one month after Caldor opened its first store in Brooklyn (DSN 4 September 1995, 61). Other major markets with dense population and land assembly difficulties have seen a willingness on the part of discounters to try urban formats. In San Francisco's South of Market District, Costco opened its first ever inner-city store with an opening day sales record for the chain. The project covers an entire city block and a four—deck garage (DSN 18 October 1993, 3). And T.J. Maxx opened its 500th unit with its first downtown store in Denver and a new downtown locational strategy for future growth opportunities (DM February 1994, 20). Clover discount store located in Downtown Philadelphia in a two- level upscale unit at 130,000 square feet at Market East mall with 170 stores . This is the 27th unit and the second largest unit, located in the vacant Gimbels (also Stems) store. The store will be open 61 hours per week compared to 90 in a typical suburban mall (DSN 21 August 1995, l). . "Several analysts say some of the best opportunities for growth (sic in the Chicago metro area) are available in the neighborhoods of the city of Chicago. While it may be more difficult to acquire the necessary land, the benefits of high population densities and consumers interested in quality retailers make the city's neighborhoods attractive (SCW September 1991, 56)." Joseph Freed & Associates, Chicago, are involved in numerous Chicago projects and specialize in redevelopment of older community centers. They pursued the redevelopment of a former Wieboldt's department store, a 10-acre site in an economically weak and underserved area in the heart of Chicago's inner city and the West Town Business District, called West Town Center. Freed purchased the land in the early eighties and proposed a two-stage project. First phase was historical restoration of the building and then 205,000 square feet of additional retail space for major tenants. The project received Cook County's first tax abatement for a commercial redevelopment. The project is now anchored by Kmart 38 and Jewel Food Store, being one of the largest centers in the city (SCW November 1991, 66). Babson College's survey of Chicago indicates a major shift in consumer patronage with the incumbent stores susceptible to new market penetration. Specifically, the survey finds that Chicago is understored, but that Kmart's 51 stores are smaller and placed closer together, effectively becoming neighborhood convenience stores for commodities, a role that used to be played by Woolworth and Kresge (DSN 5 September 1994, 10). For redevelopment opportunities in built-up areas, environmental pollution can be a key deterrent. "The problem is clean land,” says Michael Sei gel, a retail specialist with San Jose-based Blickman—Turkus brokerage. He says most potential "bi g box" locations in central cities are former industrial sites with toxic waste problems (SCW April 1994, 42). Given the scale of land necessary for major retail development, many abandoned former industrial, warehouse, and railroad yard sites are being reviewednbut environmental risk and clean-up are frequently substantial. Dress Barn's chief executive says many retailers have ducked the low income of big cities as too tough to do business in. But as new immigrants come into the population and settle into these areas, they have needs and purchasing power and he believes discounters should look again, if not at the inner city, then at areas very close to the inner city as places to do business (DM December 1993, 82). Even in light of these positive trends, the fear of crime (either real or perceived) is an issue affecting central city retailing. The riots in the sixties had a devastating effect on investment in some areas and the recent Los Angeles riots have also had a negative impact on central city investment perceptions. In the South Central Los Angeles riots, 1,360 of the 2,411 stores in the area were burned or damaged with many not returning (DSN 16 March 1992). Wal-Mart, which requires about 15 acres for its new prototypes, will have to negotiate with about five existing businesses to assemble a site in Los Angeles. As the 1i 105 39 nation's least advanced community for modern discount retailing, discounters have been hindered by expensive real estate, the sheer number of stores to serve the market, and the nation's most expensive media costing more than 40% greater than other major markets (DSN 5 December 1994, 78). Wal-Mart, which entered the California market just five years ago, plans major expansion over the next five years with opening more stores than in any other state. They now operate 79 Wal-Marts and 26 Sam's Clubs in California running a distant third to Kmart with 193 stores and Target at 115. Many of the company's new stores will be located in the expensive and land-scarce Ios Angeles and Orange counties where Wal-Mart has virtually no presence. No stores exist in Los Angeles but the retailer is moving in closer to the city (DSN 7 August 1995, 3). Wal-Mart is presently invading Los Angeles with six units in Los Angeles County, although San Francisco is still not a major presence (DSN 15 January 1996, 1). It should be noted that most of these positive trends are happening in extremely expensive, dense market locations and may not be replicable in many situations, especially where land is generally available inexpensively either in the city or the suburbs. SugrcentersuFood and General Merchandise Combined Another key concern for central cities is the impact of the new supercenter or hyperrnarket discount format that combines food stores with general merchandise in 200,000 square foot stores. These centers require at least twenty to thirty acres and will be extremely competitive for older, and even modern foodstores and discount store units. In 1960, approximately 14% of all discount stores had supermarkets under the same roof (generally operated as a leased department). In 1962, it reached 32%. Total foods sales in discounters were then $1.5 billion or 2.9% of the total grocery store sales. Management Ventures estimates that the figure will rise to 5% by 1996 given the new thrust to supercenter development (DM April 1994, 6). 40 The format is not new since conceptually a strip center with food stores and discormter department stores adjacent to each other were recognized for their complementary drawing power. Kmart had substantial grocery departments in their early stores. Fred Meyer of Portland and Meijer's of Grand Rapids have been operating these centers for several decades. And combination supermarkets attempted to capture a significant amount of general merchandise sales and other ancillary convenience goods and services such as bakeries, butcher shops, drugs, and florists in the last fifteen years. The wholesale club industry's success at attracting food sales has stimulated the rethinking of the discounters towards this approach. Discount Merchandiser defines a supercenter as a combination of a complete, f 1111- line discount store with a supermarket. The units generally range from 100,000 to 200,000 square feet, with hyperrnarkets ranging from 200,000 to 300,000 square feet. Hypennarkets include a complete supermarket as well as specialty departments such as a bakery, deli, and other services. Most have an extensive general merchandise offering, but a narrower presentation than discount stores, focusing on quick-moving items. Combination supermarkets cover at least 30,000 square feet with at least 25% devoted to general merchandise. Kmart opened its first supercenter in 1991 in Medina Ohio, generating $66 million in sales replacing an older and smaller traditional Kmart doing approximately $13 million annually. The store is 155,000 square feet and other prototypes are being tried including 190,000 square feet. By the end of 1994, Kmart will have over 70 supercenters in operation and Wal-Mart will have 140 supercenters. Wal-Mart has four prototypes at 116, 136, 167, and 188 thousand square feet. The current 165-unit supercenter division of Wal-Mart will end the year with about 240 with the future potential storecount of 1500, approximately the same number as Wal- Mart had traditional stores in 1990 (DSN 19 June 1995, 90). Wal-Mart opens its first mini-supercenter of 100,000 square feet in Cameron, MO. (6500 population) replacing a 41 20year old 47,000 square foot store. Grocery represents about 20,000 feet with fresh produce cut by 50% compared to larger supercenters (DSN 4 September 1995, 1). Wal-Mart favors their traditional small towns and county seats whereas Kmart is trying everywhere and they are convinced the market potential exists in a variety of settings. An urban Kmart tends to have higher average transactions than their normal stores at two to three times a week, with some people coming from 40 miles away. Stores are usually open 24 hours. Every Kmart supercenter has been a new facility but several of the 55 openings in 1994 were conversions. Gene D. Hoffman, formerly chairman of Super Valu and consultant to Kmart, indicates that the concept began with "people and employees focusing on delivering customer-defined values. Store associates are encouraged to get involved with churches, schools, the Scouts and other positive elements of the trading community. We empowered them." Meijer's scrapped its $100 million investment in SourceClub (warehouse membership club to compete with Pace and Sam's Club) after one year closing all seven stores and halting construction of four of five to open in 1994. They have shifted to developing supercenters in Michigan and Ohio where Super Kmart's expanding and they are opening eight centers in Indiana. Meijer‘s had 76 stores in 1993 with six new stores in Ohio for a state total of 3. They have been penetrating the Cincinnati and Cleveland markets since 1994. Meijer's has plenty of real estate buying land contingent on obtaining planning approval and necessary rezoning as far as five years in advance. Discount Store News estimates that 1993 sales would be $5.4 billion at $75 million per store. They recently obtained approval for a 215,000 square foot prototype in 1995 for 35 acres in Strongsville, a Cleveland suburb. Also, it is developing a 230,000 square foot prototype for Grand Rapids at a 55-acre site compared to an average 35. It is making a major thrust to having restaurants in its stores including leasing out to McDonald's, Pizza Pan, and 42 others (DSN 3 January 1994, 1). At the end of 1995, Meijers had 99 stores averaging 210,000 square feet. Fred Meyer has 94 multi-departmental stores that are free-standing and averaging 133,000 square feet with a one-stop shopping emphasis in 1994. Seventy carry food, the firm's new major focus. The food store portion is typieally the same size as free-standing super food store competitors (DM January 1994, 14). In total, Wal-Mart, Kmart, Meijer, and Fred Meyer have 265 stores doing about $12 billion in 1993. But the supermarket industry is $365 billion. By 1996, Management Venture believes that 40% of the $45 billion supercenter industry will be food, or about 5% of the entire food industry. Peter Monash, a Columbus OH. based consultant believes that by the year 2000 Wal-Mart will be the nation's largest food retailer with more than 500 supercenters. He states that "supercenters, hypermarkets, are simply strip centers that have architecturally been folded into a box under one roof.” Super K had 22 combination food and general merchandise stores in 1993 with plans for another 50in 1994. Approximately 133 are in various stages of planning by the major discounters. Retail rrrarketing consultant James M. Degen of Santa Barbara says that sales should total $16.4 billion in 1994 with a total 458 units, up from 305 at the end of 1993. DSN estimates that 648 exist at the end of 1995. Bernard Scsnick, retail analyst for Oppenheimer & Co., suggests that 3,600 units could be built as retailing's greatest growth concept. This concept helps bring in customers and one-stop shopping desires given consumer demand for convenience. George Rosenbaum of Leo J. Shapiro & Associates of Chicago suggests that supercenters will be the 'nub" of a power network of stores in a market by each retail player (DSN 7 March 1994, 3). As an indicator of the consumer draw of these facilities, 64% of adults shopped at a Meijer in the last 30 days compared with 38% at Target and substantially less at Kmart in Grand Rapids. In Columbus OH., 40% of Meijer's shoppers visited four times a month. 43 Kroger believes that Wal-Mart cannot be stopped and will eliminate the lowest ranking supermarkets in a community. Of Kroger's 1,280 stores, about 70% are combination stores with only new stores as combinations averaging 50,000 square feet. Food retailers are impressed by Wal-Mart's technology and logistics--the ability to put food on the shelves at lower costs than food chains. Bill Bishop, Barrington IL. consultant says that ”supercenters represent state-of-the-art distribution, compared to the traditional superrnarket's, which was built, thought about, and implemented 20 or 30 years ago.” A recent study for the supermarket industry showed that supermarkets' average prices and average costs were out of line with the three major competitors--warehouse clubs, mass merchants, and deep discount drug stores. Discounters want food because they can double or triple their general merchandise sales operating on low food margins, selling enough general merchandise to create a profit at a higher return on investment. It is not unusual for the weekly customer count in a supercenter to rise from 25,000 visits to 50,000 visits per week A key issue is to have proper food distribution, especially regarding perishables, where some weakness have occurred to date. Quality and freshness must be preserved for this concept to work successfully, in addition to price and convenience (DM APRIL 1994, 38). The discount store industry has grown rapidly and has become a major force in retailing and potentially crippling to central city retail economies. The discount store sector must also be considered in light of the simultaneous growth of the shopping center industry. S 'n Ce ter vel ment Sta The evolution of the shopping center should first be put into the perspective of the growth of department stores that evolved from the dry goods store during the Civil War period. Department stores grew with the expansion of cities and improved transit to the downtown area. And the general rise in the standard of living enhanced department store 44 formation. The expansion of the electrical age and the evolution of corporate structures with mechanization, mass production of clothing and products, advertising, and availability of capital spurred department store growth. Also, modern building techniques allowed for people to get to upper floors in the vertically-oriented downtown districts. "A department store is a departmentalized retail establishment that sells many lines of merchandise including men's and women's clothing and accessories, piece goods, small wares and home furnishings. In 1948, the Bureau of the Census added the further qualification that a department store would be required to have twenty-five employees (Smith 1956)." In 1929, 4,221 department stores in the nation accounted for 9.0% of total retail sales declining to 2,761 stores in 1954, accounting for only 7.3% of total sales. By 1958 a substantial increase to 3,157 stores accounted for 7.5% of sales. In 1987, there were 4,243 conventional department stores, but 5,798 discount stores also existed that were virtually unknown in 1950, thereby providing additional impetus for shopping center development (U.S. Department of Commerce 1989). Related to the growth of mass merchandisers has been the simultaneous growth of retail chain organizations in which firms with four or more units accounted for 23.7% of total sales in 1954, increasing to 47.8% in 1982. Thus, in concert with franchising, substantial expansion of corporate control over retail stores during the shopping center growth period occurred thereby virtually negating the significance of small, independent operations. Recent studies indicate that corporate growth--in concert with the demographic trends, availability of capital by lenders in the post-war era, and local governments' need for diversifying their tax base-stimulated the unbelievable growth of shopping centers (Ghosh and McLafferty 1991). What is a shopping center? Virtually all studies use the Urban Land Institute definition: "A group of architecturally unified commercial establishments built on a site which is planned, developed, owned, and managed as an operating unit related in its 45 location, size, and type of shops to the trade area that the unit serves. The unit provides on-site parking in definite relationship to the types and total size of the stores (ULI 1977)." The scale and pace of shopping center growth as an urban form is remarkable. The number of retail centers expanded from around only 100 in 1950 to 4,500 in 1960; 12,500 in 1970; 22,500 in 1980 (O'Neill 1991); and 40,000 today. At the end of 1993, 39,963 shopping centers existed in the United States. Representing 4.8 billion square feet of gross leasable area with over 850,000 tenants, these centers account for $ 814.1 billion in sales, or 56% of the nation's non-automotive retail sales. See Table 9 for the distribution of centers by size and sales. These facts seem staggering since this real estate concept really did not take hold until only forty years ago with most space in suburban growth areas. Open centers (i.e. not enclosed malls) comprise 70.1 % of the total GLA of all shopping centers in the United States. In 1993, 667 new centers with a total of 92.2 million square feet opened with 415 less than 100,001 square feet and 144 between 100,001 and 200,000 square feet. In 1986, 28,496 centers existed with 3.5 billion of GLA (gross leasable area)-significant growth is still occurring in spite of center saturation. Table 9 Naicawide ShoppiugCeaten. 1993 _ Gro- Square Feet E Total GLA Total Sales Average Center Average Sales Per (Billions of Square (Billions S) Size Sq. R. Feet ) (Sq. PL) (3) «00.001 24.993 1.214 n66 48.574 194.91 111M131 -200.000 9.61 1 1.312 206.9 136,552 157.65 200,001-400,000 3.166 0.84 124.2 28.448 148.96 moor-800.000 1.194 0.666 105.5 557.694 158.51 ”fill-1.000.000 295 0.265 52.0 899,485 19588 >r.ooo.ooo 374 0.479 88.9 1.31.870 185.63 Total 39.633 4.771 814.1 111,373 170.67 Source: National Research Bureau Shopping Center Databaae And Statistical Model. 1994. NotcTheNadoudRueathureaaCemhuenimmnotaaauditof ingeenterauofDecernberBl. 1993. Publisher oftheShoppiugCeuter ' forcver35yeara,itiauaedthroughouttlre' uatryandiaalaopubliahedauauallyiuthe StatiniealAbatractoftheU ' States. Stat: Ame: \‘olauf British Proud ”We 46 Many of these centers increasingly have discount store anchors especially with the evolution of the power center concept. Howard Green (1986), a retail market research and expansion planning consultant in Troy, MI. states that today's value-conscious buyers want things their way and are unwilling to settle for anything less. Discount stores keep attracting more shoppers, rising by more than 10% since 1974. Discounters have “operating efficiencies due to extensive use of technology, emphasis on low prices, wide selection of commodity merchandise, and attention to customer service. Traditional department stores, on the other hand, have experienced a 14% decline in shopping frequency since 1974." Only Dayton Hudson, Dillards, and the May Company are financially sound national chains. As an institution, the shopping center--together with the interstate highway system built in the 1950s and 1960s--has helped mold the economic, cultural and social life of the post-World War II era. Having acknowledged the enormous role of the shopping center in our lives, let us examine current trends. First, while gross leasable area per capita (GLA/capita) has risen from 9 square feet in 1974 to 19 square feet in 1992, during the same period sales per square foot have gone down from $175 to $160. Second, there has been a significant decline in both the number of trips to shopping centers since 1980 and the number of stores visited per trip. In fact, trips to regional malls are declining almost as rapidly as visits to downtown areas. A reporter recently asked me what would happen to regional malls by the year 2020. Without hesitation, probably foolishly, I replied that half of them would disappear. I will modify my fast answer by suggesting that to be successful, malls must possess superior regional access, densely populated trade areas and economically viable anchors. The others will maintain themselves as is, or decline, ultimately being worth no more than the land on which they sit (DSN 18 April 1994, 68). Rogers (1990) reviewed the status of the shopping center industry in the United States as background for international expansion by British and European companies into America, which he says has been difficult since they do not recognize the strength and volatility of competition produced by the relative absence of land use controls. Many British and European retailers are accustomed to the comfortable and profitable security provided by urban planners who prevent new suburban retail developments in their eagerness to protect established town and city centres (Rogers 1990). in 19 Be! rm; retai Oven Wfic lei 47 Market saturation has occurred due to lack of planning controls, tax shelters, and inadequate site research. Rogers claims that about 23% of shopping centres had a negative cash flow in 1990. According to Rogers, department store vacancies in the U.S. will be a greater problem than in Britain because of the relative absence of land use controls and the resulting strong competition. He suggests that for political and property cost reasons, Britain will never witness the scale of suburban shopping center development that continues in the U.S., nor the resulting intensity and commercial blight. Lord (1988) indicated that in Britain, a shopping center larger than 100,000 square feet exists for every 190,000 people whereas the ratio is one center for every 29,000 people in the U.S. Between 1972 and 1984 population and personal income grew by 13% and 44% reapectively. However, the number of planned shopping centers expanded by 86% with retail floorspace increasing from 7.9 square feet in 1972 to 14.3 per capita in 1984, creating overstoring. The American retail landscape should continue to undergo massive structural change, especially as consumers increasingly demand reasonable prices. Isadore Barmash states that only the best will survive retail saturation. Retail space was 12 square feet per capita in 1980 with 20 today. Saturation is the point at which sales of the nonleading companies begin to decline (SC‘I' October 1995, 84). Standard productivity measures can identify problems with the shopping center where a concept operates such as declining comparable store sales, sales per square foot, and traffic levels. Other comparisons can include the historical performance of the subject site, performance of other stores in the chain, or performance of other competitors in the market (CSAE August 1993). Table 10 highlights the vast increase of space per capita developed since 1987 and the significant difference in rate of center growth versus welines in sales per square foot. Table 10 Shopping Center Productivity 1%7 1%8 1%9 1990 1991 1992 Annual Growth Rate (‘5) OLA sq. ft. per 153 161 17.0 17.6 181 18.8 4.2 capita :fisfion- 162 159 156 151 144 142 -2.6 per sq. ft. (5) Sauce: International Council of Shopping Centers and Management Horizons Division of Price-Waterhouse R. Fulton Macdonald, president of International Business Development Corp. indicates that retail square footage is estimated to average about 120 percent of real demand and considered as much as 150 percent in some markets, thereby offering consumers numerous alternatives for shopping and product gratification (SCW August, 1993, 15). Yet, even with these negative aggregate figures, the shopping center industry, capital, and retail tenants are planning for major expansions. In 1987, two rrrillion retail firms existed with 496,000 having two or more locations. Theae multi-unit establishments are increasingly capturing retail sales from 33.5% of total sales in 1958 to 56.5% in 1987. At the same time that per capita sales are increasing, sales as a percent of income has declined considerably from 55.5% in 1958 to 40.9% in 1987. Per capita retail sales have grown in constant per capita sales (1982 dollars) from $3,423 in 1963 to $5,722 in 1987. Retailing is a major employer. Service sector growth has risen from 58.5% of total employment in 1946 to 75% in 1986, with retailing relatively steady at around 24% of total service employment, slightly declining since 1946 (Stemlieb and Hughes 1988). Retailing accounts for about one in every five jobs nationally. In 1990, 1.5 million retail establishments with payroll had 19.8 nrillion employees and $1.8 trillion of sales. The general merchandise standard industrial classification included 36,600 stores with 2.1 million employees and $216.5 billion of sales. Department stores (a sub-category of in 105 Sale V6131 49 general merchandise including conventional, national chain, and discount department stores) represents 10,100 units with 1.7 million employees and $166.2 billion in sales. Multi-unit department store establishments represent $163.2 billion in sales—virtually all of department store sales (U.S. Census Bureau, Statistical Abstract, 1993). Discounters accounted for 43% of publicly-held company retail sales in 1992, up from their 41% share in 1991, coming at the expense of other retailers. Soft-line specialty stores' share of sales dropped from 18.6% in 1991 to 18.4 % in 1992 while the share of sales realized by department stores dropped from 28.0 % the previous year to 24.9 % in 1992. Hard line specialty stores, however, still managed to bring up their share of sales to 13.8% from 13.3% in 1991. The Kurt Salmon Associates' report studied sales figures from public companies in the retail, apparel, textile, and footwear markets. Twelve of the 105 public retailers included in the study accounted for 68% of $138.8 billion in sales. Sales for all retail sectors increawd in 1992 with discounters leading the pack at a 17% gain versus department stores at 5%, soft-line stores at 5% and hard-line stores at 15%. The discounter trend is impacted by a change in consumer attitude from product orientation to value orientation (LUD 26, 9, 1993). Retailers in Shopping Center World's survey expect to open 30% more new stores in each of the next four years than they did in 1993. The more than 200 surveyed retailers opened 6,191 stores in 1993 and expect to open 30,490 by the end of 1997. In 1993, each retailer opened an average of 27 new stores with a projected average of 33 annually through 1997. Twenty-five percent may also increase store size or launch new chains as an alternative approach and they plan to remodel 12,000 stores over the next four years (LUD 27, 2, 1994). To illustrate the importance of shopping center development to the capital and developer markets, fifty percent of the total space under construction in 1992 by the top 20 developers listed in the National Real Estate Investor was in retail projects with nine reporting over 80% of their projects as retail (LUD 26, 3, 1993). As stated previously, 50 667 new shopping centers were opened in 1993 for 92.2 million square feet of new space on the market. The top 100 owners of all shopping centers range from 67.2 million to 2.9 million square feet with a median of 8.5 million feet (SCW January 1994, 26). Although a significant amount, the top 100 owners represent only 17.8% of the total space showing the diverse localized markets within the shopping center industry. A cursory view of the headquarters of these firms also shows broad coverage of firms throughout the United States. Sixteen of the top firms have over 20 million square feet each. The status of both discount store and shopping centers indicates the major growth still occurring in this sector. The next two chapters of the literature review provide the theoretical market related perspectives regarding these trends, in order to establish the framework for preparing an hypothesized model. CHAPTER3 LITERATURE REVIEW-~PART II THE SPATIAL AND CENTRAL CITY DEVELOPMENT PERSPECI'I VES Sfl'al Investment Strategy Location decision-making for discount stores represents a major financial commitment for a relatively long-term. As a real estate and operating investment, decisions are difficult to reverse and costly, including the opportunity costs of not being in successful locations. Furthermore, poorly performing sites may be costly from a reuse potential since structures may not be easily adapted to alternative uses and the site's potential for large scale retail use may be prohibitive based on the experiences of the unsuccessful retailer. In concert with effective mamgement, a positive retail location provides the ability to generate adequate profit and return on investment. Higher risks are associated with change ranging from unit management, merchandising format, advertising, or a new store location. Change increases the risk of failure because future predictability diminishes. "The significance of a location strategy is to minimize retail failure, or conversely, to help assure a successful undertaking (Mercurio 1984)." Also, the trend towards fewer but larger stores, offering a wider range of merchandise, an extension of self-service and longer Operating hours, has increased the capital investment in building a new outlet (Simmons 1984). Furthermore, "since store locations are long-term investments, the criteria for evaluating alternative location plans must include performance over the entire planning horizon (Ghosh and McLafferty 1987)." Berry, Parr, et. al. (1988) describe the four interdependent decisions in the store location process: 1. Selection of a market area in which new stores are to be located 51 im at! low: line subur Vane} (Latin largerc Hiram diSm'butr'r lhemselve Lang, ldenrjfimm 52 2. Identification of feasible sites for new stores based upon land availability and land-use regulations 3. Choosing the sites most likely to optimize the company's performance 4. Determining the optimal sizes of the stores. Laulajainen (1987) studied the eariy spatial history of diffusion within the retail industry including discount stores. He states that the logical strategy for a chain store is to allocate resources in order to maximize their net present value. This assumes one can forecast market potential and competitive actions in order to develop an opu'mum portfolio. Information is difficult to assemble because openings in spatial coverage and market penetration strategy might be obvious, especially for larger firms. And knowledge of store openings and age would also help in anticipating future life-cycle issues. Laulajainen indicates that access to such information is not easily obtainable with research departments at the major firms considering it proprietary. Discounters focus on low mark-up due to high turnover, extensive self service, and lower quality real estate. Merchandise initially was in hard goods but has evolved to soft lines as a major thrust. "The discounting breakthrough is commonly attributed to the suburban sprawl of urban agglomerations." Discounting evolved from scratch, through variety store chains, through department store companies, and through food chains (Laulajainen 1987). The history of retailing is the history of small business evolving into larger corporate formats. Discounters have tended to exhibit either a rural or urban agglomeration strategy (Laulajainen 1987). A balance occurs with the necessity to have adequate warehousing and distribution to efficiently handle store units. He documents the notion that rural stores lend themselves to expansion by concentric circles and contagion versus hierarchical diffusion. Large agglornerations can allow for distribution, advertising cost, and customer identification with as numerous units as possible. Many stores follow a pattern of regional diffusion until becoming national. These firms have taken a very aggressive posture ICE ICE ROI Wh. 53 towards the use of technology with Wal-Mart leading the industry. Fmthermore, these firms have shown an ability to seek out new opportunities in an increasingly segmented economy. Wal-Mart's penetration of rural America is an obvious success story of distribution where no one believed an adequate market prevailed. Furthermore, these firms are now internationalizing literally throughout the world. Taylor (1980) suggested that business decisions by primary corporations affected the spatial trends for their industries. Low markup dictates the spatial rule of keeping operations compact to get economies as stated above. Laulajainen believes that a discounter should put stores in rapid succession for market penetration. This was the strategy Target pursued in suburban Chicago opening ten new stores in 1993. But this could lead to overstoring and competitor reaction. In fact, Venture, a 104-unit chain with sales of $1.9 billion immediately remodeled four more stores in its most important market with 19 of 38 Chicago area stores now offering the dynamic elements of the 1991 prototype (DSN 21 March 1994, 3). When Korvette leapfrogged into other markets, it lost economies of scale and never recovered. Contagious diffusion was followed more than hierarchical innovation spread throughout the early discount store history (Laulajainen 1987). Deciding between entering a new market or expansion in an existing market require different risks. New market entry allows for a new well-planned strategy to be developed, but the great risk is lack of experience and knowledge of the new market including corporate analogs. While expansion is less risky, older obsolete units can constrain locational decisions to have an effective regional network of stores-~what apparently has happened to Kmart. The national holding company evolution has changed the investment structure of retailing whereby firms are relatively footloose and capable of investing capital in any chosen region. Thus, their location flexibility provides concern for those who support local control of the community's economic base (Bluestone et. al. 1981). Access to scale economies by larger companies generates additional capital, enabling firms to expand lea doc. "filly arebar and Cl: 54 geographically, grow internally, or diversify outside the industry. Being footloose could be especially troublesome now for communities with older, obsolete Kmart units. Given their scale, the large discount organization has an easier ability to reallocate resources between facilities to enhance overall profitability. Alternatives exist on-site or through locational shifts in facility operations, even internationally. Berrnan and Evans (1989) express the importance that ”retailers who invest large capital expenditures in land, buildings, equipment, and so on often have return on investment (ROI) as a company objective. Retum on investment is the relationship between company profits and investment in capital items." Increased efficiency in operations is an objective of many retailers. Due to the rising costs of land, building and construction, and associated interest rates, the significance of avoiding poor or mediocre locations has heightened (Mercurio 1984). Store location is affected primarily by competitors, transportation access, population density, type of neighborhood, and neamess to suppliers. The terms of tenancy are vital (rent, flexibility and term) and whether to build, buy, or rent. The physical attributes, or atmosphere, of a store and its surrounding area greatly influence the consumer's perception of the retailer. Size and characteristics of the surrounding population, level of competition, access to transportation, availability of parking, attributes of nearby stores, property costs, length of agreement, population trends, legal restrictions, and other factors all affect spatial decisions for retailers (Berman and Evans 1989). Retail location decisions are really a combined investment consideration of both the shopping center developer and the retail tenant. Shopping center developers play an important role in guiding and limiting the locational choices available to retailers (Rogers 1990). And retail tenants, especially anchor stores such as discount department stores, really guide the developer decision to invest through long-term leases by quality tenants that are bankable as collateral for financing. Some discounters prefer to develop their own sites and even their own shopping centers. Many prefer to lease space in the belief that the real 19! bulll~u 3180 is 00de 55 estate investment requirements for ownership are neither oriented to their experience nor their investment requirements. Whereas the following discussion focuses on the development community, the same conceptual strategy approaches are being considered from the discounter on the operations side. Investors seek both psychological and financial rewardsuthe psychological enjoyment of a revitalization project in a city can be immeasurable, but financial benefit must exist. The ability to have a showcase piece of real estate, the positive press locally and even nationally, the admiration of neighborhood residents and local government officials, and the pursuit of social objectives frequently involved with city developments (e. g. construction and permanent hiring goals, job training and creation, aesthetics, public enhancements, etc.) are psychological benefits to many potential investors. And yet the return on investment in the center must surpass the investor's possible yields on alternative investment opportunities. Investors can be either owners and/or lenders. Generally, the investor is interested in the interest retum on mortgaged funds, the nature of the collateral, and some form of equity participation. The yield on investment comes from a combination of these sources (Hines 1988): 1. Cash flow 2. Tax-free proceeds from refinancing 3. Equity build-up through mortgage repayment 4. Capital appreciation 5. Tax shelter 6. Hedge against inflation One can understand the risks involved regarding these sources when considering a built-up or declining area versus a new growth area. Cash flow is not just the base rent but also is determined on a percentage of sales or overage so that the developer must be confident in the short and long-term market potential of the site to attract retail shoppers and 1'. [1; dc H101 im’oh'e dim: 00 56 sales. Also, expenses must be analyzed regarding risks of vacancies, debt loss, repair and maintenance, security, and other fixed and variable costs compared to other center investment alternatives. All of these sources nwd to be evaluated in relation to the life cycle concept, not just of the center, but of the surrounding trade area. The Risks Of Investment A. The Planning and Construction Phase These risks include finding the best location, the most advantageous method of financing the land, the correct market assessment for potential sales, and the judgment about raising capital. The construction risks are apparent but the transaction costs of getting local approvals and possibly local government assistance can put city centers at a disadvantage. Typically, suburban sites especially at the peripheral growth stage of a community require rezoning and capital improvement issues with relatively limited public involvement. But in-fill developments in existing built-up areas create a more extensive time dymmic for approval processes and public exposure-financially and personally as described later in Barriers to Redevelopment. B. The Operating Phase The primary risk is that the projections for revenue and expenses are met. Again, more risks at a city location exist especially over the long-run given life-cycle considerations of an existing mature trade area versus a new growth area. C. The Termination or Sale Phase The transaction costs and yields for future sale of a central city project need to be evaluated. What will be the investment climate at the projected time of sale and will financing be available for the project at that site? The fact that a store is potentially profitable is not sufficient to generate a recommendation or a commitment for a new site. New business undertakings always involve an element of risk. Return on invested capital should be at least equal to what the client could earn in a federally insured investment, although sometimes the client will settle for. Obs ‘an 50pm enmur 57 for less, since growth is an inteng need for most businesses (Berry, Parr, et. al. 1988). Since shopping centers in a central city are deemed higher risk (perceived or real), are the returns on investment higher than other alternative investments where risk is deemed less? Most central city developments are a higher risk given the lack of firm market knowledge and trade area life cycle issues. From a market equilibrium perspective, capital and labor have moved to high return on investment areas thereby generally leaving the older central cities behind. Strategic decisions by discounters emphasize long-term policy goals and objectives, operating perspectives of scheduling, budgeting, and distribution, and administrative decision-making lines of authority and feedback. The evolution to some decentralization of management through micro-marketing is a substantive change in prowdure. Thus, production and market location theories need to be modified by behavioral organizational frameworks. Factors Affecting Retail Development Trends Stemlieb and Hughes (1991) highlight three factors that stimulated the growth of retail development in the 1980s. Wealth effect: Surging home values in the 1980s created equity growth and the opportunity for households to consume more. Borrowing Boom: The 1980s was a great decade of debt through the ease of credit and the ability to borrow on home equity loans. Demographics: The era represented the baby-boom's evolution into household formation, first-time homeownership, and peak consumption years. This generation has observed the growth of the shopping center and the consumption society, with mass consumption given their rapid growth in household and total income. The decade also saw the rising standards of consumption with the acquisition of increasingly higher priwd, sophisticated, and designer-labeled goods. International economic competition also affected industrial restructuring and encouraged investors to retreat to office and retail construction. Federal tax policy, 58 especially the 1981 Economic Recovery Tax Act and federal deregulation, encouraged tax driven construction, especially in retailing that historically had been protected by inflation. "Retail facilities were erected in order to serve the new jobs, residences, and households generated by a boom economy, although the latter part of the decade saw a decline partly due to the 1986 federal tax laws.“ Many markets became overstored in the eighties and may slow down retail development opportunities. In 1989, retail sales in the U.S. surpassed $1.7 trillion. From 1981 through 1988, retail sales grew by 58.9%. Rogers (1990) believes the 19808 development was assisted by generally low land cost, relative absence of land use and planning controls, and the tax shelters afforded by the pre-1986 income tax laws. Also, "...the role of the baby boom cannot be discounted. Baby-boomers were born in suburbia, raised in suburbia, settled in suburbia, and now work in suburbia, tales of gentrification and yuppies notwithstanding. Given the dominance of this suburban generation, it is not surprising that its employers would locate accordingly. The suburban growth corridor is intimately linked to the maturation of the baby boom (Stemlieb and Hughes 1988)." Another critical aspect of the evolution and expansion of suburban retailing regards wholesaling, ”the growing reliance on our critical trade pipelines--the suburban highway networks-and the economic reality of expansive single-floor structural configurations-- versus the multi-floored parcels of older land-poor areasuhas pushed the physical handling of merchandise to the suburban ring, a phenomenon that shows no signs of abating." Wal- Mart, for example, has 27 distribution centers throughout the nation. "It took the development community nearly a generation to fully awaken to the possibilities of the new one-hundred percent locationsuthe metropolitan circumferential highways and their radial corridor outgrowths. It has taken only a half -dozen years to build well past the market saturation level. ' may 300d define and on SbOppe 511181312: are W1 suburbs, l shopping : 59 Retail trends are not solely consumer-driven since they are also propelled by technological change, new product categories, and new approaches to merchandising (Rogers 1990). Technology enhancements are deemed the key reason for growth and potential growth opportunities in the future including anticipated direct linkages on sales between retailers and manufacturers. Stemlieb and Hughes indicate that consumers are insecure in the 1990s due to deteriorating shifts in income dynamics, potential tax increases, the inverse wealth effect due to housing value welines, and reduction in debt burden. Given the maturing baby boomers with their peak income potential, they also have peaking expenses with children entering college and the necessity to save for retirement. Although these trends create apparent negative conditions for retailing in general, the rise in consumer orientation to value shopping places the discount industry in an excellent position for market share penetration increases, perhaps at the expense of the weaker regional malls and department stores that are challenged at the higher end by specialty stores. Central Cities Vs. Suburban Trends Although some central cities have improved their retail base since the mid-1970s with major downtown redevelopment projects, these developments have primarily slowed the pace of decline but have not reversed the loss of retail sales, especially in the shoppers goods categories (Breheny and Hall 1989). Shoppers goods (GAFO Merchandise) are defined as general merchandise (i.e. departrnent/discountlvariety stores), apparel, furniture, and other miscellaneous shoppers goods (i.e. books, stationery, sporting goods, etc.). Shoppers goods sales are especially important to central cities because they generate substantial sales and taxes from both city residents and shoppers not living within the trade area. With large-scale regional malls and new power center formats being built in the suburbs, many central city residents have altered their consumer behavior patterns from shopping in the downtowns and older neighborhoods to the suburbs. Over 5,000 f ull-line 800 cent (10% lhesh Muller Want beflerbe ”ions (1 60 discount stores exist nationally, primarily located in the suburbs and now surpassing the number of department stores. Various studies have shown the severe negative impact of shopping centers upon central cities and their downtowns. Of a sample of 159 metropolitan areas, retail sales declined by 23% in the CBD with significant losses in all regions of the country, were stagnant in the central city, and had gains of 19% in the suburbs from 1972 to 1977 (Morrill 1982). Kellerman (1985) indicated that central cities had 51.4% of the population in 1960 and 58.5% of sales in 1958; but by 1970, cities had 47.2% of the population and only 38.8% of total retail sales in 1972. By 1962, major regional centers-as defined by the U.S. Census-surpassed central business district sales (Kellerman 1985). And yet these negative central city trends cannot be ftrlly explained by the movement of population and income (Muller 1981). In his study of ten large northern cities during the 1972-1977 period, Muller found that shoppers goods sales declined by 24.6% while gross income only declined by 7.6% and non-shoppers goods only declined by 10%. Muller's findings suggest that, while cities have been able to retain convenience goods and services since they tend to be neighborhood-based, cities have lost a substantial portion of their shoppers goods sales because of the substantial development of shopping centers outside of the central cities. Much of the literature has focused on macro-trends for retailing, especially comparing downtowns to metropolitan area trends. Whereas these studies have indicated that most of the shift in retail sales to the suburbs is a function of income and population growth trends (Baerwald 1989; Breheny and Hall 1989); Kellerman 1985 and 1988; and Morrill 1982), Muller (1981) indicates that more sales losses have occurred in central cities than would be warranted by their related population/income declines. Southern and western cities fared better because of the ease of annexation based on more positive state statutes in these regions (Breheny and Hall 1989; Morrill 1982; and Muller 1981). 61 Breheny and Hall (1989) cite the reasons for decentralization including increased auto usage, city congestion, economics of trade from larger units, and the movement of population. Location factors shaping spatial patterns of retailing also include capital availability, sales potential, the logistics of service and supply, and the elasticity of profitability with respect to sales potential (Craig, Ghosh, and McLafferty 1984). They also indicate that land availability and zoning, optimization of site selection, and optimal size and design considerations are paramount to store location decisions. Furthermore, they suggest that stores locate in a metropolitan area in a neighborhood wave pattern with additional outlets being located at increasing distances from the initial outlet. Thus, city locations might be a reasonable location for developments as the suburban area becomes absorbed. Baerwald ( 1989) suggests the vital importance of being located on premiere metropolitan transportation networks, and he identifies older major regional centers that were doing better than others primarily due to the transportation variable. Bradbury, Downs, and Small (1982) indicate that central cities have done well (i.e. fiscally sound, limited problems, etc.) when they are in warm climates, have MSA employment growth with the city capturing its share, have a low percentage of old housing in the city limits, and have a relatively low percentage of minorities. Bourne (1991) cites the following that hurt the central city position: * global restructuring " multi-nodal suburban development * large-scale development companies Downs adds these reinforcing negative traits: * racial and income disparities with suburbs * older housing with limited maintenance “ property taxes higher in the city * weaker educational system due to the above * lack of funds for rehab and development 62 "' lack of land availability * lack of adequate market support for retail and services * changing agglomeration and scale economies for facility locations Ghosh and McLafferty (1991) indicate that the movement of retailers from the cities to the suburbs coincided with the diffusion of planned shopping centers, the rapid expansion of retail chains, the shift from local to corporate ownership of stores, and increased concentration in retailing. They indicate that the changes in retail format and corporate ownership easily facilitated the shift from one location to the other. Furthermore, they indicate the importance of access to capital in the financial markets after World War II as a key factor in stimulating development and the growth of large-scale shopping center developers. As the developers and major anchors become partners in concert with capital availability, the shift to the suburbs was accelerated. The growing scale and concentration of retail development has moved the locus of initiative from the individual business to the large-scale developer and the retail chain, aided by their own or private market-research groups (Berry, Parr, et. al. 1988). Regarding the future, several researchers believe that older suburbs and central cities will be potential sites for development because they have been passed over. Breheney and Hall (1989) believe that the long-continued process of outward deconcentration of America's urban areas is over and some scholars identify certain counter-forces that are causing a real reversalnurban revitalization, gentrification, and city service sector growth (Leven 1988). Lord (1985) indicates the suburban retail landscape has been saturated such that downtowns, infill sites, and middle-markets are the next development frontiers. Certainly, retail development trends have occurred swiftly over the last several decades. Forecasting the potential for central city retailing is a difficult endeavor, given the rapid changes in the retailing climate, but it is incumbent upon central city policy-makers to be aware of the constraints and opportunities in their communities in order to capitalize on retail development potentials to strengthen their tax base and provide job opportunities. su‘ IDCI pred over ] m'fletir metro}; by 0111) “01331“ househok 63 The historic monocentric meuopolis does not exist today as a new threshold of suburban development has been reached through an infrastructure and land use network. Consumption dominates the American economic scene with mass marketers and dominant retail giants scrambling market share and new growth opportunities. Many losers are anticipated in the evolution of new distribution centers. Stemlieb and Hughes (1988) believe the central city's market share will continue to shrink across-the-board; it will be just one of a number of economic constellations within the new extended metropolis. Joel Garreau (1991 and 1994) has made the case very well in his edge city concept. He shows that only 22 of America's 40 biggest job centers are in downtown areas and thoroughly displays the continuing economic marginalization of central cities in metropolitan America today. Kasarda (1988) shows that the urban centrifugal shift in the last 75 years remains with the suburban ring gains coming at the expense of both the central cities and non metropolitan areas. "From 1975 to 1980, 5.7 million more central city residents moved to the suburban rings than vice versa. During the 1980-1985 interval, the net central city- suburban migration exchange favored the suburban rings by 6.2 million, or an additional 500,000 over the previous five years.” Both the "back to the city" movement and the non- metropolitan revival of the late seventies were overshadowed substantially by more predominant shifts to suburban rings . And during the last decade, the number of people living in the suburbs increased by over 18% or almost twice as much as the total population has grown. Similar trends in the nineties are anticipated (Berman and Evans 1989). While the number of jobs in major metropolitan area suburbs increased 159% between 1960 and 1990, central city jobs grew by only 24% (LUD 26,6, 1993). The median age of the American population advanced from 30.0 years in 1980 and is projected at 34.3 today. Married couples with children have declined from 40% of all households in 1970 to 27% in 1988--a major shift in consumer household-types; over 24% 64 of all households are now made up of just one person with a continuing decline of household size to around 2.6 persons per household compared to 3.14 in 1970 (Rogers 1990). Regarding some key future trends, Ms. Norma Rusbar, Ernst & Young Real Estate Consulting Group, San Francisco highlighted demographic facts pertaining to shopping centers (SCW October 1993, 148). * During the past 10 years, the U.S. population grew 10.5 % but growth is projected at only 7.7% from 1990 to 2000. * Per capita income growth grew in real dollars by 20.9% in the same period with only 2% projected for this decade. * As the population ages, it moves from the household formation stage to the established stage where they need less goods. Expenditure growth is expected in those homes headed by customers in the 35-54 age group. * Industry rule of thumb is that each person can support between 12-14 square feet of shopping center space. In 1980, 13 square feet existed and 17.7 by 1990. If no more retail space were built until 2000, 16.4 square feet of space would exist to support an estimated demand of 12.8 square feet per person, a major oversupply of aggregate space. "‘ Storeless retailing in both television and direct mail may create 7% less demand. * Yet, recent Census Bureau revisions indicate that the 1990s will jump to 25 million population net increase versus an estimated 17 million and projections upped by 17 million to a net increase of 56 million in 2010 (DM December 1993, 6). A major announcement on December 4, 1992 has far-reaching implications for the central city and urbanization in the future. The U.S. Census Bureau announced a revision of the pepulation forecast significantly upward over the next 58 years with an increase of 50% to 383 million by 2050-an increase of over 83 million above prior projections (Harper 1992). And a substantially higher percentage of minority population is projected. Although the article did not project household growth, approximately 100 million housing 65 units presently exist-so, a minimum of 50 million net new units will be built just to meet household growth. Obviously, related land use needs for other activities such as retailing will be required. What are the implications of this forecast on the urban spatial pattern, especially for suburban growth and central city revitalization? Given all of these trends, Breheny and Hall (1989) identify three various growth periods over the last forty years when the shopping center evolved and they predict a fourth generation of centers to be in-fill development. Landis (1989), however, identifies severe constraints for central cities in the future with increasing growth of jobs in the suburbs, the baby-boom generation leaving the central cities for the suburbs, and the increasing concentration of minorities and low income groups in the central cities. Even if centers located in the cities, it would be imperative that they be strong enough to attract city shoppers. Consumers rate location alternatives on their basis of the total utility of the stores and not merely their location. Such aspects as store image, appearance, price, service levels, number of checkout counters, employee composition, availability of credit, and parking determine the ability of a store or shopping center to attract customers (Craig, Ghosh, and McLafferty 1984). Furthermore, studies indicate that between 30 to 50% of all shopping trips are multipurpose thereby reflecting the necessity to have an agglomeration of facilities within a limited area in order to attract consumers (Ghosh 1986). Thus, the city resident shoppers might pass by a new city center if another location in the metropolitan area provided them a wider array of goods and services that could be acquired within one area thereby limiting consumer travel expenditures in cost and time. bong-term trends for retailing are proposed by many to be smaller-scale development with much more specialimtion and market segmentation. Given fragmentation of consumer markets, scarcity of time, computer technology and interactive video, and service worker shortages, Bates (1989) believes that localized and national specialty stores will be the wave of the future. Many analysts agree even to the point that Management Horizons (1990), a subsidiary of Price-Waterhouse, believes that there will be fewer store firms, isa fun 66 contraction of retail footage, storeless shopping, and the decline of the mass market. They even predict the demise of the discount store with a shift to commodity megamarts with a direct link between manufacturers and retailers. Other researchers debate the future of the department store given these trends to specialty markets. Rather than moving in the direction of mass marketing, Mason (1986) suggests that the new emphasis on market niches and superior service gives department stores a new life cycle and identifies Nordstrom's as an example. Shetlr (1983) believes that pluralism in our values, lifestyles, and behavior is likely to increase thereby creating shopping opportunities for minority markets. Places can and do reinvent themselves, and political decisions can shif t the scene of the ”action," but long-term shifts are not simply due to raw demographics (Stemlieb and Hughes 1988). While analysts do discuss overstoring (Levy 1989 and Management Horizons 1990), Sterrrlieb and Hughes (1981) conclude that both inner ring suburbs and older passed-by city markets may be a key thrust for the future. Kellerman (1988) demonstrates that non- downtown and non-major regional centers had the largest and increasingly growing sector, especially from 1977 to 1982. Given the rise in specialty markets and smaller centers, now is a good time for central cities to consider their market position for development in the future. Furthermore, the bulk of the regional mall infrastructure is in place for the near term, and new central city opportunities for developers and retailers may meet a demand for capital investment. However, city government, developers, and retailers are deterred from developing major retail centers in the city even when market analyses show demand. Land availability is a major problem in an existing built-up environment. Even when sites can be found in a proper location, cities typically have to use their powers of eminent domain and subsidize the cost of eliminating blight, a timely and costly process. Developers worry about the costs involved and the ability of the government to perform expeditiously. Furthermore, they are concerned about the long-term financial feasibility and return on investment in city 67 locations given the heterogeneous population and frequently changing dynamics of central city neighborhoods. While retailers have similar concerns, they are also apprehensive about the impact on their store operations and whether they need to alter their merchandising techniques. Thus, central city governments have numerous hurdles to overcome to attract major retail development. W Howard Green indicates that ”life-styles are no longer predictable from income, social class, place of residence (rural versus urban, apartment versus house), or type of employment. Life-styles appear now to be based on interests, activities, and personal needs and wants.” (Green 1988) Green says that consumers are looking for a coherent strategy at either the conventional or low-margin ends of the spectrum: kLigLE-ad 1. outstanding assortrnents 2. appropriate ambiance 3. good service 4. reasonable prices mm 1. satisfactory assortrnents 2. good in-stock positions 3. reduced transaction time Berman and Evans (1989) confirm that strategy by suggesting that consumers are more knowledgeable, selective, and affluent. And they are especially more price conscious. Tailoring stores specifically to their local trade areas is a recent effort by American retailers by serving the segmented marketplace better versus the cookie-cutter merchandising and operations that enabled chains to steal market share from independent retailers. Scrambled merchandising is now standard in many formats as retailers try to offer shoppers the 68 convenience and time-savings of one stop shopping and to attract them more frequently. Discounters see food and groceries as traffic-builders that can increase shopper exposure to their higher margin non-food merchandise lines (Rogers 1990). Two major forces have affected this scrambled merchandise trend (Rogers 1984): 1. The general escalation of operating costs has (and is) forcing all retailers to seek more rapid rates of sales growth in order to preserve their profit margins. This pressure has obviously been experienced most sharply by retailers in static businesses, such as the supermarket industry. Supermarket retailers have progressively added non-food and general merchandise lines in order to replace the food sales lost to the growth in eating out at restaurants. In the extreme case-the food—drug combination store-this process has involved the integration of an entire 'drug store' merchandise array into the supermarket format. 2. Certain retailers have recognized the competitive weaknesses of other store types in adjacent businesses, and adjusted their product mix in order to attract business away from the weakening segment. The growth of the chain drug store industry at the expense of the traditional variety store is the most obvious example of this process. Interestingly, both of these ”newer" formats now have competition from the newer discount supercenters of today. Working men and women with no children under the age of 18 living at home are the most active shoppers according to the 1992 Consumer Research Study commissioned by the International Mass Retail Association. The most active shopper tends to be a married, college-educated, professional, baby-boomer woman with no children under 18 at home. Saturday is the most popular shopping day with strip centers the most shopped with shoppers preferring to shop in one place within 20 minutes of their homes. Key motives for shopping are low price, 39%; selection, 37%; location, 33%; quality, 14%; and in- stock availability, 9% (LUD 26,1, 1993). And according to the Fall 1993 Retail Satisfaction Index compiled annually by Milford OH. based Frequency Marketing Inc., 38% of consumers named discount department stores as their outlet of choice in 1993, up from 29% in 1992. Today's consumers are cherry-picking items for the lowest cost at a variety of store types. The traditional department store has dropped considerably in men's and women's apparel with the specialty store seeing an increase. Discount stores were generally given a poorer rating 69 in the apparel category, although their sales have been steadily increasing (DM April 1994, 14). Philip Lempert of Age Wave, Inc. indicates that the average 65-year old has an annual discretionary income of over $6,000 compared to $3,200 for a 35-year old. The 50+ shopper represents the most powerful buying group in America representing 25% of the population, yet 50% of the discretionary income. In 1990, the average woman became a grandmother at the age of 46 and is buying for four grandchildren with a median age of 8 (Berman and Evans 1989; DM April 1994, 72). 'By the year 2000, the traditional primary target of U.S. marketers, the 18 to 34 year olds will shrink in numbers by more than 9 million while the number of 50+ year olds should increase by 10 million. By 2020, it is projected that another 36 million 50+ consumers will be added to the population, bringing the number of adults 50 and over to 112 million out of a total U.S. population projected to be 336 million. One-third of the population will be over 50! (DM April 1994, 69)“ ' e cle f Devel ent and rade Areas--The Cen Ci Dile a The life cycle concept suggests that products, industries, technologies, facilities, and neighborhoods have various stages of development from initial growth, through maturity, decline, and obsolescence. While some aspects for retail distribution are within the general control ofthecornpanytomaintainthegrowthandmaturityoftheoperation, tradeareas (and their related neighborhoods) are almost totally out of control of the retailer. Furthermore, the average life of areas and facilities appear to be shorter now, primarily due to rapid technological change. Various models have been deveIOped on the life cycle concept Based conceptually on the von Thunen land rent theory of the early 1800s, land use models have been created by sociologists, economists, and geographers primarily on a monocentric concept with concentric circles of land uses and socioeconomic groupings modified by transportation networks and secondary market nodes throughout the metro area. The well-known 70 theories of Burgess, Hoyt, Harris and Ullman, Vance, Hoover and Vernon, Birch, Lowery and others are descriptive of United States metropolitan spatial land use patterns over especially the last century (Hartshorn 1992; Winsberg 1989). But virtually no attention has been given to dealing with a long-term condition of urban decline and the possibility that equilibrium conditions will be difficult or impossible to establish (Lang 1982). Various models have been formulated based on urban density, age of physical plant, business life cycle, community power, and neighborhood life cycles. Law (1988) suggests that the rate of suburbanization in metropolitan areas varies according to the following factors: 1. Attractiveness of the central city 2. Accessibility of the central city 3. The structure of office activities in the region 4. The rent and local taxes gradient between the city versus suburbs 5. Preferences of the population and lifestyles 6. Cost of transportation 7. Character, attractiveness, land availability, and land price of suburbs vs.city 8. Rate of metro growth 9. Degree of political f ragrnentation and related competition to city for residential bundle of services 10. The strength and objectives of the planning system Given these factors and the rate of decentralization, the monocentric city is really obsolete with the intensity of activities reaching their peak in the downtown and declining evenly towards the suburbs. The metropolis of the future will be polycentric in spatial form (Law 1988). The out-migration pattern is primarily a market verdict on older city neighborhoods, which are not delivering those elements that residents, business, and industries expect and can find elsewhere (Lang 1982). 71 Certainly, race has an important impact on neighborhoods as shown by invasion- succession theories pertaining to racial transition and eventual turnover. It is impossible to overlook racial change since so much (but certainly not all) of severe neighborhood decline has occurred in the path of expanding black ghettos (Varady 1986). Varady indicates that the tendency for ghetto expansion to lead to neighborhood decline is really attributable to income versus race as initial middle-income black families are replaced by lower income black families. The stereotype-that once racial change begins, virtually complete turnover occurs-is supported by reality with very few examples of successful stabilization efforts in racially mixed commturities adjacent to black ghetto areas. Although neighborhood decline is associated with racial change, this does not necessarily mean that the black community should be considered totally responsible for this decline. Racism is certainly a key, if not fundamental, cause of such problems as poverty and racial segregation that contribute to social class and racial change. The United States has relied primarily on the private homebuilding community to provide housing, with relatively high standards through restrictive zoning and subdivision regulations for more expensive homes, stringent building codes, and the provision of exceptional public facilities such as parks, libraries, and schools in new growth areas (Bradbury, Downs, and Small 1982). These high quality standards legally exclude the lower income population from new growth areas. Thus, relatively higher income households have historically moved to the periphery of the metropolitan area thereby creating a reliance on the filter-down process to primarily provide housing for lower income households, who move into vacated housing by those who have moved 'up" in the market. While this pattern seems to have increased after World War II, the concept of higher income households moving outward has been an historical pattern in the United States since its conception. Outward movement was assisted by continual transportation improvements, substantial land availability, capitalist investment in infrastructure and growth, and personal financial ability to "escape" the ills of the oldest portions of the 72 central city such as density, disease, smoke, noise, lack of greenery, congestion, and heterogeneity of land uses and economic groups. This filtering process creates a chain of vacancies descending to lower income groups until eventually the supply is larger than the demand and a housing unit is either reused with an alternate land use, temporarily or permanently vacated, or demolished leaving a vacant land parcel or converting to a new land use. Various studies have shown some of these chains to range from 2 to 7 units based on the value of the unit starting the chain (Adams 1984; White 1971). Furthermore, White indicates that half of all moves in the U.S. lie in chains set off by new housing while the other half is set off by deaths of families. Given the older age of residents in central cities, their death could stimulate a rapid chain reaction among lower class households. The steady relocation of effective demand for housing (and related retail goods)--away from the less desirable stock and toward the more desirable-stimulates above-average price rises in the growing, high demand areas; but the relocation of households depresses real prices in low-demand areas, which are usually areas of heavy net outrnigration (Adams 1984). Emptying-out decline in older areas normally does not occur without high levels of new suburban housing construction in relation to net household formation (Bradbury, Downs, and Small 1982). From the viewpoint of each neighborhood affected in the filtering process, change is perceived as undesirable decay. But from an area-wide perspective, this decline is a necessary part of accommodating growth of the low-income population. However, some people who would have preferred to remain in older city neighborhoods, have been injured economically by disruptive neighborhood change. Many have felt forced to move and have felt that this process has been a hardship-and it usually is financially. Commercial establishments and their markets have been directly affected by these processes. Public Affairs Counseling (1975) prepared The Dynamics of Neighborhood Change for the Department of Housing and Urban Development in anticipation of the Community Development Block Grant Program. The document was prepared to describe 73 neighborhood change processes and establish programs varying by the stage of neighborhood decline. The process of neighborhood change is triggered and fueled by individual, collective, and institutional decisions made by many people-households, lenders, real estate brokers, investors, speculators, insurers, businesses, public service providers (police, fire, schools, sanitation, building codes), and others. If one could influence these decisions, one could presumably affect the neighborhood decline process. The analysis indicates that prior efforts have failed due to the failure to see that the decline of neighborhoods is a total process with clearly definable stages; the inability to intervene early, at the preventative stage; and the lack of recognition that neighborhoods are not independent of one another within the process. Every neighborhood changes all of the time with about 20% of all households in the U.S. moving annually. In some city neighborhoods, 70% annual turnover is not uncommon. Change is not necessarily bad-~the issue is the degree and in what direction. Each neighborhood goes through a life-cycle from the time it is built to the time it is demolished. Following the life cycle concept, neighborhoods fall into five stages-- healthy, incipient decline, clearly declining, accelerating decline, and abandonment. This description of neighborhood evolution includes five basic steps along with subsequent variations and repetitions (Stemlieb et. al. 1974): 1. Racial transition--from white to black 2. Decline in average income of residents as a result of the 'filtering' proeess 3. Declining levels of security accompanying increase in number of low-income households 4. Increasing difficulty with tenants involving rent payment, maintenance of the parcel, and tumover 5. Inability of landlords to obtain loans through normal market channels The remaining steps in the process are combinations of 2 through 5 as they interact with each other to produce steadily worsening conditions: 74 6. Physical deterioration 7. Declining tenant quality 8. Psychological abandonment by the landlord 9. Final tenancy decline and evacuation-actual abandonment. Not all cities have every stage. In fact, many cities do not have accelerating decline or abandoned neighborhoods. Neighborhoods may change quickly or slowly, and may stay in one stage for a long period of time. However, unlike the human life-cycle, some neighborhoods can regenerate. Knowing the stage of decline and the trends affecting that neighborhood can greatly assist policy makers in determining programmatic responses. These processes affect the trade area for retailing, yet are beyond the control of developers or tenants. Thus, market research and site selection must anticipate the likely patterns of neighborhood change for a long-term and predict the implications for an investment. Solomon and Vandell (1982) describe very adequately the three theories of neighborhood decline, which directly affect the retail environment. The orthodox economic perspective emphasizes the concepts of markets, competition, static equilibrium, and allocation of resources by price. The process of decline originates with lowered expectations of return on investment. Tire dual theory is based on a socioeconomic calculus focusing on the persoml relationships between landlords and tenants or between residents and other actors in the housing market such as financial intermediaries. The foundation of the duality is the degree of compatibility between the landlord and the tenant and the landlord's motive in ownership. His/her motive is not primarily as a capital asset but more of a personal or household economic framework. The primary cause of neighborhood decline under the dual theory is a breakdown of the reciprocity agreement. The radical theory is neo—Marxian including landlord/tenant class conflict and economic/political power distribution. Thus, radical theorists deny competitive markets 75 contending that the low income market is a monopoly of landlord and financial interests. Radicals believe that decline is driven by capitalists to maximize their share of productive surplus in real estate by preserving existing power relationships. The importance of acknowledging the operational theory, obviously, is to devise programs that will counter the negative trends of the theory. All three theories have value and are involved in different neighborhoods of many large cities. Solomon and Vandell did not conclude which theory is most relevant. Rather, they were laying the foundation for further research. Central cities are dependent partly on the amount of suburban growth. The amount of new construction in a metro area is a function of the area's economic condition, its number of households, and the age and economic characteristics of those households (Bier 1988). It is also dependent upon adequate transportation and availability of land (i.e. topography, soils, zoning, infrastructure) at a reasonable price for development-~reasonable being defined by the local market. Investment decisions and ability to obtain financing will depend on the valuation of the property in light of the life-cycle process. How do real estate appraisers look at this aging process for properties and neighborhoods? Appraisers value real estate on the basis of one or more of the following approaches: sales or market comparability, capitalized income, or replacement cost as modified by depreciation. The value of a parcel of real estate is directly related to the property's physical, functional, and locational characteristics (Hickman, Gaines, and Ingram 1984). Properties typically have two traditional classes of obsolescence (Klaasen 1989): 1. Functional obsolescence pertaining to the service function that a structure can provide; and 2. External or economic obsolescence pertaining to factors affecting the economics of a property such as location, changing lifestyles, government rules, zoning, financing, etc. Many central city properties have both types of obsolescence and investment decisions are 76 made on the basis of whether improvements and maintenance will pay dividends through either higher income, lower expenses, longer economic life, or greater personal comfort. External obsolescence can even be considered as ”negative feasibility.” A property must meet minimum investment objectives of an investment scenario to be a feasible investment (Galleshaw 1991). Failure to meet these objectives will affect the mortgage- equity position in the reverse order of the real estate collateral by first eliminating profit, developer‘s/owner's overhead, and if declining rapidly, part or all of the equity investment and of the mortgage position as well. Another way to look at this negative property value is that real estate is an equity in which the responsibility and liability of ownership exceeds present worth (Weinstein 1973). In valuation, the total economic life at the time of appraisal depends on the quality of construction, the stability of the location, the age of the property improvements, and local market condition. The life of a building is a function of the age of the improvements. However, actual age and the effective economic life used for older improvements are not equivalent (Derbes 1987). According to The Dictiopgy of Real Estate Appraisal, external obsolescence is defined as "an element of accrued depreciation; a defect, usually incurable, caused by negative influences outside a site.” In addition to these two categories, temporary obsolescence could also occur for the short or long-term, but it is very hard to place a value. For example, garages built originally for the Model T became obsolete from the 19408 to 19603, but became appropriately sized since then for compact cars. Even many of the gentrifying urban neighborhoods are occurring in areas with small, functionally obsolete row houses originally built for low and moderate income households. But now they are well located, have charm, and most importantly, are appropriately sized for singles and dual working households without children. Obsolescence is not necessarily the result of age and is difficult to value because structures may not even be physically deteriorated. 77 Neighborhood deterioration and eventual abandonment is partially an economic process reflecting the substantial increases in the supply of urban land, thereby allowing rapid suburbanization. The question of central city land value and the economic life of central city buildings is directly affected by the outlying competition. Furthermore, older central cities have lost their economic vitality as the pattern of American industry changed in response to technological improvements, automation, and foreign competition (Weinstein 1973). Since the post-war era, Boume (1991) indicates that we have also witnessed systematic devaluation of much of the capital stock by investors, landlords, and governments through accelerated rates of depreciation and shorter horizons of profitability. Simply stated, the real estate market and resulting abandonment of certain areas, including an avoidance by the shopping center industry, is merely an application of microeconomic theory with supply and demand factors (Anderson and Funderburk 1989). Appraisers look at four broad forces in valuation—social, political, physical/environmental, and economic. Anderson and Funderburk identify the following appraisal principles within the microeconomic framework: 1. Anticipation—value exists on the basis of future benefits in spite of sunk costs. 2. Change—market conditions are not constant and are affected by consumer preferences, investor expectations, and other factors 3. Increasing and decreasing returns—the concept of diminishing marginal utility pertaining to real estate needs 4. Conformity-property values of an area are maximized when relatively high degree of architectural homogeneity and compatibility of land use exists. This factor suggests that heterogeneity and lack of conformity exerts a negative influence on central city values. 5. Balance-the relationship between an entire property and its environment is crucial in determining returns on investment and, therefore, value 78 6. Substitution-consumers tend to pursue the lowest priced good when similar benefits or utility are offered These valuation factors provide the rationale for the difficulty of declining neighborhoods and properties to retain their value. In effect, these factors highlight the prisoner's dilemma for investment by various parties. The prisoner's dilemma theory indicates that two adjacent property owners are limited in their investment decision-making by what the other will do (Hartshorn, 1992). If one owner rehabilitates or redevelops and the other does not, the rehabilitated/redeveloped property will be negatively affected by the adjacent deteriorated property and will not receive as good a return on investment as the one who did nothing. Thus, both property owners are in a dilemma regarding their investments. If both improve their properties, they will both theoretically receive good retums, benefiting from each other's investment. This example is just for two properties; the dilemma is more complex for property owners on one block or for a whole neighborhood. The importance of mutual decision-making” perhaps induced by the local government-is required to improve a neighborhood's values and long-term future. Through urban renewal and community development programs, cities have attempted to eliminate these negative extenralities affecting value. The shopping center developer and retail tenants making long-term commitments are integrally tied to the above life cycle concepts and issues. Making a commitment for twenty plus years has been more easily understood on greenfield sites where all new development occurs in a relatively homogeneous economic orientation that should have a prolonged successful life cycle. Both the life cycle and the prisoner‘s dilemma concept point to the necessary role for the public sector to attempt to eliminate negative extermlities in central city environs. Why Central Cities Need Discount Stores--The Pro Growth Persgtivg As stated previously, every city is a different market and every site has its own impacts. City taxation and programmatic requirements vary; and city needs and priorities 79 vary. Each situation requires its own benefit-cost and impact analysis for potential developments. However, the following is the general pro-growth discussion of the potential importance of discount stores to central cities. Discount store retailing has been described herein as the major retail force in American shopping, especially for the last twenty years. Although some planners would prefer smaller-scale retail establishments, others believe the market has opted for large-scale efficient distribution centers that have quality, convenience, wide selection, and low prices representing value. Cities should continue to encourage small scale retailing in select agglomerations. However, some planners believe that neglecting the discounter will only continue the retail decline of the city. The discounter could serve as the anchor for other community retail strategies on otherwise obsolete commercial strips by generating traffic. If shoppers already leave the city to go to suburban discounters, then some planners suggest that the city should encourage discount store development to create an attraction for other retail opportunities including generating new retail incubators for minority and small business enterprises. Discount stores are a major employment force in the nation employing many lesser skilled individuals on both a full and part-time basis. Discounters have approximately one full time equivalent employee per 600 square feet of space, or 167 jobs for a 100,000 square foot unit. These are not necessarily net new jobs to retailing, however, as some retail establishments rrright close due to more efficient competition. Although not the highest paid positions, retailing offers an opportunity for many interested in part-time work to find employment. Retailing may provide an opportunity for unemployed and underemployed to both have a job and learn vital working skills. Since many cities must provide various means of assistance to attract discounters, most cities usually have a variety of social goal requirements that must be met such as job training; city resident, minority, or unemployed priorities for permanent employment; construction contract and job priorities 80 for small, minority, or female conuactors and residents; and small tenant assistance in shopping centers. Discounters may be important to many central cities from a tax standpoint, especially if cities have a local sales tax. Nationwide, local sales taxes represent 7.1% of local government revenue. And sales taxes represent 10.0% of the total revenue for these fifty cities, but 28.7% of the city's own revenue generating tax capacity. For example, a 100,000 square foot store generating $20 million in sales would create $200,000 in sales taxes at 1% (not necessarily net new). Also, sales taxes could create the opportunity to capture sales (and thus sales taxes) from suburbanites who only shop or work/shop in the city, as well as from tourist sales. Furthermore, many cities have various "head” taxes per employee, even increasing in tax rate with establishments having a large number of employees. And others have local income or earnings taxes whether one lives or works in the community, thereby creating another potential opportunity to tax non-city residents for jobs created in the city. Also, many communities are especially reliant on the property tax. Discounters and shopping centers are relatively large net taxpayers requiring less cost to service (e. g. fire, police, garbage, etc.) than residential neighborhoods, especially in a central city. Many centers are substantially privatized in security, maintenance, and waste removal thereby potentially reducing their costs to the community. Most shopping centers, and especially those with national tenants, are well insured, inspected for code and safety issues, are built to modern codes, and have sprinkler systems. Most real estate assessment procedures are based on the income approach, which is another potential inducement to have cities encourage the location of large income generators for property tax purposes (Berman and Evans 1989). Every city's fiscal sources of revenue vary and the economic necessity for retail development, from a tax base standpoint, will depend on the tax system of each community. Many cities are dependent on property taxes as their primary source of 81 revenue. But the city of St. Louis, for example, only receives 9% of its income from property taxes whereas local retail sales tax creates 11%; an earnings tax for people who either work or live in the city representing 32%; and business employer taxes of 11%. On the other hand, Michigan cities relied substantially on the property tax until 1994 and no local sales tax existed-the state received sales taxes and shared part of the revenue with communities on a per capita basis regardless of the level of sales tax generation in the locality. Whereas retail development might have been beneficial for a Michigan community for other reasons than retail sales taxes, certainly the perspective of local government about projects they encourage and the impact on all taxing bodies would be influenced by the tax consequences of the individual state and locality. Unlike some other land uses, shopping centers frequently create ancillary opportunities for spin-off developments including retailing, office, and housing. A discount center could create additional opportunities to clean-up adjacent commercial blighted and non-blighted areas either through rehabilitation or redevelopment. Neighborhood stability, preservation, and revitalization can also be enhanced by having attractive retail settings nearby. City shoppers also are value conscious and want convenient, modern, value oriented shopping opportunities. Various surveys have shown that residential location decisions are made partly on the availability and convenience of retail shopping opportunities. In addition to providing a retail offering, city shopping centers can elirrrinate a blighted area and enhance adjacent residential property values. A major shopping center investment can also be the catalyst or positive sign of reinvestment and commitment to an area. The psychological value of such an investment might also spur the private sector real estate and financial community as well as local government to continue their efforts in such neighborhoods. Positive psychological values are directly beneficial to residential property owners by enhancing the perception of appraisers, Realtors, lenders, insurers, and investors in the area. Neighborhood organizations could be stimulated to be even more active. 82 City officials are concerned about the potential negative impact of new shopping facilities on existing business. Businesses in the immediate area may become more organized or stimulated to encourage continued upgrading and investment. While some businesses may be unable to compete with the new retailers, others could improve their efficiency by enhancing their quality and selection of merchandise, improve their technology, enhance their service, improve their interior and exterior appearance, upgrade their marketing, and capitalize on the increasing awareness by the consumer of the area as a place to shop. Although major discounters attract many shoppers, nearby businesses could look at the attraction as an opportunity to capture additional sales, especially new customers who never shopped in the area. While some of the existing shoppers at smaller neighborhood stores may be attracted to the new discounter, many of the city's discounter customers may have already been shopping at discount stores, but many miles away in the suburbs. Thus, the city's position could be that it would be able to retain discount sales from residents within the community (and all the tax benefit) versus providing income to the suburbs. Some large cities have attempted successfully to enhance downtown retailing through festival marketplaces, enclosed regional-style malls, mixed-use development with a retail component, rehabilitation/facade restoration programs, and various attempts at marketing and tenant leasing endeavors. Some of these developments have been successful, at least temporarily, in attracting downtown worker shopping, tourist/convention shopping, and city and suburban residents. These centers have been an important cog in many downtown development programs for expanding convention/hotel facilities, offices, stadiums/arenas, cultural facilities, tourist attractions, mass transit, parking facilities, other retail opportunities, and residential redevelopment and preservation in or adjacent to downtown. While each market varies and downtown is a key priority in many cities because of the tax base, some cities cannot effectively compete by encouraging a major downtown retailing program given the local market parameters. However, most cities can support Per hen WK (an i 5‘ al. t‘pec. 83 some level of discount store facilities in the right locations. Presently, one discount store (includes both full line and limited discount stores) exists for approximately every 10,000 households in the United States. Even those communities with strong downtown retailing should have the market opportunity to increase city retail capture by providing discount retailing in the city's neighborhoods. Local Government Revenue Sources In 1991, all local governments raised $410.3 billion in revenue from its own sources with sales and gross receipts representing $32.0 billion, or $736 per capita (U.S. Bureau of the Census, Government Finances, No.5). City governments collected $19.6 billion or $127 per capita in sales/gross receipts taxes, which represented 27.1% of their local tax revenue raising revenue. Property taxes generated $37.7 billion at 52.2% of city tax revenues. Table 11 reflects the importance of all revenue sources for municipalities and especially the importance of property and sales taxes to their own revenue raising ability. Table 12 identifies the importance of various municipal taxes based on population size with sales taxes being an important locally-generated source of income. Although sales taxes are important in the aggregate, a review of the 50 central cities being analyzed show major differences in its importance ranging from 0.0% to 82.6% as a percentage of local taxes with a median of 30.3%. The property tax as a percent of local taxes varies from 7.7% to 100.0% with a median of 51.3%. So, each community's tax base differs significantly and the importance of retail development from a direct tax revenue perspective may (should) influence local decision-makers in their development planning benefit-cost evaluation. Taxes can certainly have an inter-urban and intra-urban effect on locational patterns, although such information is beyond the scope of this analysis at the micro-level. Taxes can have an effect at the corporate level regarding investment decision-making (Bluestone et. al. 1981) on a regional basis. And rate differentials in communities can have an effect, especially those taxes directly paid by the retailer-usually property taxes and 84 business/income taxes. But sales tax rate differences within a metropolitan area could eliminate certain sites from consideration. These factors also pertain especially to restructuring and decisions pertaining to reinvestment actions. Table 11 General Revenue For All Municipalities. 1990-91 Item - 1990-1991 Percent Distribution Percent Of Local Taxes General revenue 164.319 100.0 Inter-governmental 46.260 28.2 From state 34.901 212 From federal 7.615 4.6 From local 3.744 2.3 General revenue from own sources 118.059 71.8 Municipal taxes 72.213 43.9 100.0 Property 37.654 22.9 52.1 General sales 11,738 7.1 16.3 Selective sales 7.866 4.8 10.9 Income 9.595 5.8 13.3 Other 5.359 3.3 7.4 Charges and miscellaneous 45.846 27 9 Current charges 27.221 16.6 Sewerage 9.308 5.7 Hecp'uals 4,035 2.5 Interest earnings 11.355 6.9 Special men 904 0.6 Sale of property 528 0.3 Other 5.88 3.6 Source: U.S. Department chommerce. Bureau ofthe Ceuus. City WFW: 199091. GFI91-4. June 1993. Table 12 thGovernrnentfinaneeByPopuluionSize. 1990-91 (Local Tax Sources As A Pementage Of Municipal Tax Revenue) M Populdicn Municipal All 1010.01!) 500.000- 300.000- 200.000- 100.000- 75.000- less Than taxes Cities Or More 999,999 499,999 299,999 199,999 99.999 75.000 Property 52.1 42.7 518 44.1 52.5 61.5 64.2 59.2 General 163 14.7 12.4 24.2 14.9 17.7 183 16.6 sales Selective 10.9 10.9 10.2 14.0 13.2 10.9 8.9 10.3 sales Income 133 25.1 17.5 9.2 9.7 2.7 2.4 6.0 Other 7.4 6.5 8.1 8.4 9.7 7.0 6.1 7.9 Source: U.S. Department of Commerce, Bureau of the Census. City Govermnt Finances: 1990-I991, GFI91-4,Jnne 1993. was To Redevelpgpent Land assembly at a reasonable cost in a good market location is frequently a major deterrent in a central city or built-up environment. Approximately 10 acres is necessary for every 100,000 square feet of retail development with surface parking. Given the desire to have an agglomeration of retail facilities either on-site and/or at adjacent sites, the required land for major discount and shopping center retailing is substantial. With the new power shopping center format in the 400,000 to 800,000 square foot range, 4080 acres of well located, accessible land are required. Very few central cities have such land available within their corporate boundaries or, if available, the land is frequently surrounded by blighted neighborhoods or industrial areas. Even a single discount department store and supermarket or a new supercenter requires 20 acres of land. Usually, sites need to be somewhat square requiring adequate frontage and depth for building location, parking, and circulation. An example of a land assemblage in a built-up area of a central city is described to illustrate the complexities. Assume the city wants to create the opportunity for a 200,000 square foot shopping center, including a 105,000 foot discount department store, a 65,000 foot supermarket, and 30,000 feet of ancillary retail space plus a couple of outparcels for 86 fast food or a drive-up bank. Such a development would require around 20-23 acres that is generally square with at least 800' depth. Assume a city block with a 15' alley is 500' x 350', approximately 4 acres. Assume that street rights-of—way are 50'. In order to obtain a roughly square site, a site of 1050' x 965' could be assembled by acquiring five city blocks and their public rights of way—four full blocks and two half blocks to the alley at the rear of the site. Under this scenario, 178,250 sq. ft. of the total 1,013,250 sq. ft., or 17.6% of the 23.3 acres, would be in public rights-of-way that could be sold or possibly donated under a development plan. Assume that the front two blocks only have commercial structures with depth to the alley and the remainder of these blocks in parking (frequently, this parking portion could be residential with a moderate density of 50—100 units)-- approximately 125,000 sq. f t.(first story only with occupants; additional second story underused for another 125,000 sq. ft. of structure) of existing commercial buildings in varying conditions may be on the site with potentially numerous owners/tenants. The remaining three blocks may have anywhere from 100-200 residential units in owner/tenant occupancy at densities of 8- 16 units per acre, not extremely dense for an urban environment. What might the cost of assembly be in a major, moderately-priced, older city (e. g. not a San Francisco or New York; rather a Detroit, St. Louis, Cincinnati, etc)? Table 13 highlights the potential costs of this land assemblage scenario. Developers might typically pay from $3 to 5 per square foot for such land in standard suburban sites. Assuming $5 is the market value, then the assembly cost of $9 suggests the need for a land writedown (grant) of $4 per square foot, or approximately $4 million for this project. Using St. Louis taxes for calculations, this type of development would create approximately 325 jobs and $600,000 of annual general revenue for the city (not all net new). To appreciate the scale of this cost, a $4 million subsidy would represent 16% of the St. Louis's annual community development block grant for housing and economic development activities, primarily for the benefit of low and moderate income persons. Table 13 Potential Land Amembly Cod For 200.000 Sq. Ft. Shopping Center In Built-Up Urban Area Activiy Sq. Ft. Structure Con/Sq. FL Total Cost (3) (3) I . . . Residential 2111.01!) 15 3.000.000 Commercial 250.0(1) 10 2,500.01!) R I 60. ( I 0 325'“ 5 1.67.5.“ Occupied Demolitionlasbectcsl 450.01!) 2 9001!!) environmental Utility and street 1 1010.01!) demolition/relocation Total 10.81 For Private Land 9,025.01!) 8.91 (Call $9) If Public R.O.W. Donated Note:CostsDoNot1ncludePublicAdministrationCoctsForC‘ PerformanceAndlegachrkFcrErm’nentDomsinOr Relocation;RelocationF1timateIsModest—CouldBeEven$2 'lionHigherBasedOnNatureOfOccnpancyAndComparable HousingCoctsForRelocateec. CostsForTwo-Story Buildings Without Major Structural/EnvironmentalProblerm. In order to avoid these major assembly problems, many city officials and developers look for large-scale, underdeveloped sites under one or several owners-they are hard to find. Greenberg and Popper (1994) describe potential development opportunities in TOADs, temporarily obsolete abandoned derelict sites. Such sites may be abandoned housing developments, factories, warehouses, schools, dump sites, railroad lines, canals, parking lots, military bases/defense plants, or waterfronts either in private ownership or in public ownership through the tax foreclosure process. Occasionally, the site is an abandoned or virtually obsolete retail development with an obsolete multi-story department store and related mall shops, requiring either clearance or major redevelopment. Sometimes, these sites are previous LULUs (locally unwanted land uses). Although Greenberg and Popper suggest that many of these sites are potentially valuable inner-city land, they are frequently surrounded by blight or have costs of environmental clean-up, rehabilitation, or clearance/redevelopment that are unfeasible without major land writedown and public assistance. 88 Some developers have even proposed building on the comer of neighborhood or community parks in order to avoid the land assembly problem. Many parks are bounded by collector streets or major arterials and are well located for shopping accessibility. Based ontheneighborhoodunitconcept, manyparksarecentral tolocal tradeareasandare actually frequently good locations for a shopping center; however, the politics of commercial use of parkland is difficult to imagine even if some attempt is made to replace the space elsewhere. Possibly, this concept might work best under a major redevelopment scenario where the park could be adequately replaced although the historical and neighborhood sentiment toward the park might be difficult to ignore politically. Barriers to land use succession and redevelopment exist and the role of the public sector in redevelopment has been determined primarily due to these barriers that inhibit the market to respond freely. The public purpose of the role of government in eliminating blight and encouraging redevelopment has been supported since the Supreme Court's Berman v. Parker decision in 1954. The primary barriers to redevelopment focus on eliminating negative exterrralities including: 1. Assembly of a large enough area to encourage redevelopment and eliminate negative neighborhood spillover effects. While some neighborhood activists (Stegman 1979) believe this assembly can be done piecemeal by parcel or by block, many economists and planners believe that large acreage must be assembled or “controlled" varying in degree by the impact of adjacent areas for complete revitalization and recreation of a growth or stability cycle. Conuol means the ability to eliminate spot blight and encourage property maintenance and investment including code enforcement and rehabilitation in order to elirrrinate the problems of the "prisoner dilemma" Programs to deal with any negative social aspects including police protection would need to be established on a targeted area basis. Thus, the shopping center developer may not only have the problem of assembling the development site, but may need to have the general area under ”control" to eliminate negative neighborhood extemalities. 89 2. Hold-outs who will not sell or who ask truly exorbitant prices deter redevelopment. 3. Ability to get clear title on old central city property is frequently difficult, due to clouded titles based on age of records, and finding/negotiating with present owners may be impossible. 4. The sheer volume of acquisition transactions with risk of failure in concert with opportunity costs for time and effort expended. 5. Neighborhood activism and political representation that may not be supportive can stop development at any time in the process. Most analysts believe that much more neighborhood activism and media review and scrutiny occur in city developments and more bureaucratic red-tape exists increasing time for project review and approvals compared to standard suburban deals. 6. Hidden costs beyond the norm occur in the central city especially pertaining to environmental hazards such as asbestos, PCBs, underground tanks, lead paint, and toxic/hazardous waste. 7. If intending to do infill projects or complementary projects with other property owners, they may not perform expeditiously or according to required plans, thereby negatively affecting one's improvements. 8. Compared to standard suburban site opportunities, central cities tend to have more red tape and include many socially worthwhile programs, but at a potential disadvantage for the city developer versus the suburban location. Such programs include linkage, minority/city resident construction and permanent employment opportunities, permanent job refenal coordination for the project's life, Davis-Bacon construction cost standards if using federal funds, and others (Nyden and Wiewel 1991). Frequently, design standards for city projects are extensive compared to the typical suburban development for enhanced facades, landscaping, lighting, parking, construction and code requirements. Bidding processes and public works coordination may be more difficult. Even noise and dust 90 mitigation during construction in a built-up environment may be more costly than a suburban site. 9. The costs of land assembly (costs of the property, required relocation, environmental clean-up, demolition, and site preparation) are usually higher than the value of the site for the reuse intended. Obviously, the private sector has a potentially difficult time in pursuing a redevelopment project because of the inherent risks involved-except for a few smaller projects where the return on investment may be adequate such as immediately adjacent to downtown. Many of these issues (especially 1, 2, 3, 4, 7, and 9) were covered under the federal urban renewal program that existed from 1949 to 1974. Cities still pursue such endeavors primarily through the Community Development Block Grant Program and use of the state statutes and local ordinances for urban renewal. The two primary tools to assist redevelopment are the use of eminent domain by the city (or developer as agent) and the provision of the land writedown, to be paid by the government to eliminate the social cost of blight to society and get land back on the market. Thus, the differential between assembly cost and reuse value can be subsidized by the government to elirninative negative externalities. land writedown has been a key role for the government since the 1949 Housing Act with the start of the urban renewal program. The additional approach of using either tax abatement or tax increment is another tool that can be used separately, but frequently is used with eminent domain and land writedown. Eminent domain assists in eliminating hold-out problems, eliminating title problems, and acquiring properties at fair market value. While these barriers are extensive, local governments that use these tools and substantially reduce the transaction cost burden to the developer can make a real difference to encouraging redevelopment. Whether the city should be involved in encouraging redevelopment has been debated from various perspectives . 91 Urban Politig Based on these barriers and the identified role of the public sector to eliminate negative extemalities, central city retail development in older areas depends also on the local political regime and its perspectives toward encouraging development. City investment is frequently dependent upon public assistance in land assembly, public financing, and coordination with local governmental bureaucracies. Thus, the local political situation and the legal, financial, and administrative capacities of the city become a vital link in the development process. Several theories exist pertaining to these roles of local government and the regimes supportive of development ranging from neo-classic to Marxist theory. Prime considerations of redevelopment policy focus on efficiency versus equity. Tensions exist between private market forces, centralization, scale economics, and investment returns versus political demands, resource allocations, and equity considerations—in the light of budget constraints. Optional programmatic responses for communities include assisting supply, demand, or human capital for economic development. Thus, spatial development strategy (i.e. by location or by type of land use) is an important part of public policy given the limited financial capacity. These considerations are all in the context of the political environment that is either pluralist, elitist, or regime oriented. Peterson (1981) looks at land development from the perspective of communities seeking highest and best use, maximum utilization, and fiscal health with a unitary perspective of public policy that development automatically benefits the city. Peterson sees the city leaving a unitary interest as a corporate entity with organizations as instruments for the efficient pursuit of goals. Furthermore, he sees the role of the city being developmental rather than redistributive in its efforts, leaving the role of equity and justice for the federal government (Kantor and Savitch 1993). 92 Stone (1989 and 1991) and progressives see more of a concern about equity considerations and a class struggle at the local level pertaining to development policy. Their primary concerns focus on who benefits by public policy. They suggest that development policy should not just be economic alone-rather, it should consider fairness of costs/benefits, justice, and the opportunity to participate in decision-making. Dahl (1961) believes that development policy tends to be pluralist in nature, even though social and economic notables do tend to be more involved in redevelopment. He believes that leadership is the key and that there are three types of people who influence policy--leaders, sub leaders, and constituents. From the political economic context, neo-Marxists observe development policy as capital seeking accumulation, exploitation of labor and natural resources as commodities, and using the state to secure capital's accumulation. Fainstein (1991) indicates that only protests and progressive movements can alter the capitalist course. Fainstein and Fainstein (1983) indicate that the city is not a unitary political community, but rather a site for class and racial conflict as expressed by the form of the built environment. Market forces produce cumulative advantages for higher income groups and continuing disadvantages for lower economic groups. While Banfield cites market forces as a given and Peterson suggests that federal policy should be geared to redistribution, David Harvey (1973) suggests that revolutionary action is required. Harvey says the single most important factor determining the redevelopment potential of a city is its position in the national system of cities in order to receive capital flow. The regime paradigm differs from the structuralist Marxist view and Peterson's view of utility maximization. Given all these perspectives, Stone brings in the importance of politics and actions of community actors in somewhat of a structuralist framework. He presents the concept of the Social Production Model whereby the regime consists of diverse actors who have institutional access to get things done. Regime theory tries to link the community power structure and urban political economy. Local governing coalitions vary 93 and help shape patterns of urban development. His main concern is to broaden the horizons of decision-makers to look holistically at decisions, who benefits, who and how participation is to occur, how is policy presented and negotiated, and how public policy is perceived. DiGaetano and Klemanski (1993) describe how regimes vary based on whether the political orientation is market or government driven; whether a progressive, activist, or caretaker government; whether in a growth management or growth inducing posture. Kantor and Savitch (1993) describe the private sector as producing “economic resources that are necessary for the well-being of the political community-including jobs, revenues for public programs, and political support that is likely to flow to public authorities from popular satisfaction with economic prosperity and security. " Thus, the public sector assists in this process through market intervention. The nature of local government can influence development, but it is difficult to study a cross sectional view of the discount store sector in light of differing regimes and their capacities over time. Even the nature of local government (e. g. strong mayor, city manager, commission, at-large, or ward systems) could have influences, but probably more dependent on the people involved. Cities vary considerably in the nature of the local regime including these groups and different perspectives: 1"Planning and community development-related agencies *Quasi-public development authorities *City bureaucracy *Business community-downtown, large vs. small, industry vs. commercial, owners vs. tenants *Neighborhoods and local advocacy groups "Media *Local, regional, state, and federal resources I"Market potentials 94 *Nature of local government and leadership style *Legal and administrative limits The structural determinants of discount store space in any location, but especially in a central city, would need to consider the nature of the local political regime as an explanatory portion of a model. Accounting for this variable would require extensive qualitative research, but a partial attempt to incorporate the political factor in the model is the inclusion of sales taxes as a revenue source. The importance of sales taxes to a community could be a prevailing econorrric interest for the fiscal health of the community as a unitary interest. The conclusions will elaborate on recommended further research pertaining to the local political framework beyond the scope of this study. racism David Rusk (1993) prepared an analysis of general city versus suburban trends in the nation's metropolitan areas, describing their relationships in terms of ”elasticity.” Cities with the greatest elasticity had vacant city land to develop and the political and legal tools to annex new land. Inelastic cities are typically older cities already built out at higher than average densities in the industrial age and either unable or unwilling to expand their city limits. Elastic cities tend to be much younger in their development cycle. Zero elasticity and low elasticity cities only expanded their city area by 6% and 44% respectively in the forty-year period whereas medium to hyper-elastic cities expanded their area by 306% and 625% respectively. Most inelastic areas are in the Northeast and Midwest; most elastic areas are in the South and West. Between 1950 and 1990 more than 80% of the 522 central cities in the nation expanded their boundaries by 10% or more, expanding their land area by 164%. This concept obviously pertains to the urban political economy affecting the ability of central cities to attract modern development. Summgy This chapter focused on the important concept of the life cycle of real estate development and its effects on lmg-term positive retums on investment. Developers and tenants are looking for sites that meet potential long term operational success. Central cities have continued to decline from a retailing standpoint due to the maturation and declining life cycles of many neighborhoods and due to barriers to redevelopment, especially land assembly constraints. The urban political regime becomes even more important in central cities as required public-private partnerships to foster development become paramount. Also, the opportunity for boundary expansion through annexation has been identified as a positive attribute for selected central cities. Although central city economic and retailing decline has occuned, some analysts believe that metropolitan market equilibrium is being achieved through suburban market saturation thereby providing new opportunities for passed-over central city retail development including for discount stores. More mass retailers are focusing on micro-marketing techniques at the store level attempting to meet market niches, thereby assisting the opportunity for heterogeneous central cities to potentially attract retailers that previously focused on relatively homogeneous, new suburban environments. The next chapter focuses on specific research and variables in model formulation used to identify successful retail locations on a metropolitan and site-specific basis. CHAPTER 4 LITERATURE REVIEW-«PART III MODEL THEORETICAL FOUNDATION Retail analysts typically evaluate the opportunity for penetrating a market area on the basis of supply and demand factors. Measures of retail saturation depend on the premise that any trade area can support only a given number of stores or square feet of selling space devoted to a particular goods or service category. Any given spatial market contains a relatively fixed amount of total market potential in terms of consumer dollars (La Londe 1961). Among the ratios used to calculate saturation are the number of persons per retail establishment or square footage, average sales per store, store sales per capita or household, and sales per square foot of selling area (Berman and Evans 1989). Charles A. Ingene, a Professor in Marketing, has written a series of macro-related retail analyses of SMSAs since 1980 that serve as the background for this analysis. Several models exist to aid in the individual site selection process dependent upon an exhaustive analysis of possible sites within a community (Ingene 1984). He summarizes that "the decision to enter a large geographic market, such as a city, represents an even greater commitment of scarce resources, for most retailers will open several stores within the area in order to take advantage of scale economies in advertising and distribution. It is clearly inefficient to analyze all sites in all areas in order to choose which communities to enter. Hence, the ability to determine the market potential of each area would be valuable. Surprisingly little research has been done on this topic.” Ingene and Lusch (1980) studied market selection parameters for department stores and indicate that the supply of retail sales may be viewed as either the number of stores in the geographical market or the total square footage of those stores. They believe both 96 97 measures are reasonable and are reflective of past entry decisions in the imperfectly competitive department store industry. This analysis uses square footage; number of stores can be extremely deceiving given the major shifts today to larger store units. Certainly, a major difference exists competitively if the market is composed of two stores of 50,000 feet versus one store of 100,(I)0 feet; and now even up to 200,000 square feet per unit. For example, Kmart replaced an older unit in Medina, OH. with $13 million in sales for a new supercenter generating $66 million (DM April 1994, 38). Demand is viewed as total department store sales in the market determined by environmental variables of population, income, and other factors effectively beyond the control of the industry. Ingene and Lusch (1980) indicate that a corollary of this view is that only the relative shares of individual department stores, and not total department store sales, can be influenced by managerial actions. This concept of aggregate supply and demand has been generally accepted by retailers and academicians. Retail operations look at potential market areas for entry or expansion especially if current firms are achieving excellent financial returns. Yet, financial data on competitors is not easily available by either a market area or site-specific basis (Ingene 1984). So, firms have used proxy measures such as the Index of Retail Saturation (IRS) developed by La Londe in 1961 for this aggregate analysis computed as : POP; s EXPi RSSi where: IRS; = index of retail saturation for area i POP, = population in area i EXP; = per capita retail expenditure in area i RSSi= retail selling space (in total square feet) in areai Ferber (1958), Hoyt (1969), Liu (1970) and others showed that total retail sales in a community are a function of a small set of environmental variables. Hoyt argued that sales the hou dem mane Opera 6We “will indepfn 98 are dependent on population growth and household formation. If correct, the IRS theory provides the basis for making market entry decisions whereby market areas are attractive if the Index for Retail Saturation is high thereby showing demand is great relative to supply of retail facilities; conversely, the market is unattractive if the IRS is low. This method suggests that demand is beyond the influence of retailers and is environmentally determined (Berman and Evans 1989; Ghosh and McLafferty 1987; Ingene and Lusch 1980). Ingene and Lusch (1980) describe the inadequacies of this theory looking at 1972 department store sales per household ranging from $232 to $1,448 for 213 SMSAs of the 258 in the nation. They determined that population and income were significantly related to total department store sales accounting for 95.7% of the variation at the p = .05 level. However, they argue that , if demand is determined by population characteristics, it should be possible to account for total sales as well as per capita or per household sales in an area on the same variables. They revise the IRS slightly by looking at total department store expenditures/household in anticipation that expenditures per household can be expanded by managerial actions such as market entry or store expansion. They use a set of independent socio-economic variables in the 213 SMSA's for the analysis regressed on total department store expenditures per household and conclude that the environmental variables by themselves do little in explaining the cross-sectional variation in retail expenditures per household (Adjusted R2: .221) and that the IRS model can be misleading since household demand is not strongly determined by environmental variables. Ingene and Lusch (1980) tested an expanded model that include variables over which managers have a degree of control such as merchandise assortment, service levels, hours of operation, atrnospherics , and locations. This model showed a major increase in explanation with an Adjusted R2 = .890 demonstrating that retail expenditures per household are largely determined by the managerial decisions of department store executives. Ingene (1984) discounted this model later by suggesting the model included an independent variable relating to size already found in the dependent variable, thus den mer- 99 accounting for the high explanation. The most significant variables affecting sales per household are the per capita department store selling space in the area and the number of employees per thousand square feet of selling space (Ghosh and McLafferty 1987). They conclude that the ability to predict total department store sales with environmental variables Ins led to the erroneous belief that per-household sales are also environmentally determined. This belief suggests that solely using the IRS could preclude entry into some market areas that could be profitable. Their model would be revised to show that the total market can be expanded as well as the need to consider the probable market share, which the new entrant can capture. See Table 14. Table 14 Revised Model of Market Attractivenem Expenditures per household Mandible mm L91! not attractive po-ibly attractive IRS High probably atuactive very attractive Source: hgeneChadaAandRobutFJmeh'MarkuSebcdonDecbiomfaDepamnenStaea' deofkeariflng, 56. 309%): 39. The wide variation in department store sales per household cannot be strictly determined by environmental variables alone with managerial decisions such as merchandise assortrnents and service levels having a substantial effect. Ghosh And McLafferty (1987) state that the IRS is meaningful only if compared against some norm such as a standard set by the retailer based on past experience. Alternatively, the relative attractiveness of different market areas can be ranked using the index. Analysis of the competitive structure is vital. The number of existing stores, the size distribution of the stores, the rate of new store openings, the strengths and weaknesses of 100 existing stores, short and long-run trends, and the level of saturation should be evaluated in relation to an area's population size and growth, not just in absolute terms. Population density is also an important factor (Berman and Evans 1989). From a planning perspective, the IRS serves as a model of aggregate equilibrium for total retail sales. Yet, retail sales are composed of many categories of goods that can be provided in a variety of evolving store types, especially given the growth of scrambled merchandising. Although the absolute amount of consumer expenditures for any segment of retailing is relatively fixed within any given trade area, it is entirely possible that a trading area might be unsaturated in one segment of retailing (e. g. drug stores) and over- saturated in another segment of retailing (food retailing). The planning policy question focuses on the role of planning to encourage continued growth opportunities and modernization of space, while simultaneously providing balanwd growth. Most planning studies on encouraging commercial growth look strictly at a balanced saturation or equilibrium, which includes functionally and economically obsolete space being a part of the existing supply. This creates what could be termed as the 'Wal-Mart dilemma" Growth advocates would suggest that Wal-Mart provides modern, efficient retailing providing quality and breadth of merchandise, low prices, convenience, and consumer satisfaction whereas it might cannibalize older, obsolete stores in a trade area or even small, well-run operations. A high level of intertype competition exists in the retail industry with the overlapping and scrambling of merchandise (Ghosh and McLafferty 1987). This ”survival of the fittest" concept runs somewhat counter to standard academic planning models that tend to encourage a supply that is in balance with environmental factors. thimm Network/Portfolio Store location strategy determines the spatial pattern of outlets that best meets corporate goals and objectives. To select optimal sites, a balance is achieved between the 101 requirements of corporate marketing objectives and the needs of the marketplace with the availability and desirability of individual sites. Opening new stores is risky with rising costs of real estate and construction, in addition to the operations cost of establishing a store. The firm's image requires successful store locations, in addition to the firm's necessity to generate sales and profits. The sales potential of a retail outlet depends on the quality and price of the merchandise, its physical characteristics, the characteristics of customers, the level of competition, and the relative accessibility of competing stores. Modeling the variability of sales potential at different locations is key to location analysis. While current sales and profits are important, sole reliance on these may be limiting with the necessity to look at both immediate and long-term retum. The entire planning horizon must be considered given the nature of long-term investments by shopping center developers and tenants. Marketing environments constantly change including changing consumer preferences, shifts in residential patterns and related demographics, and the effects of the competition's decisions. Ghosh and McLafferty (1987) indicate that ”the forces of growth and decline, gentrification, and neighborhood abandonment make up the social landscape to which retailers must respond and which determines in part the viability of retail institutions.” Increasing geographic coverage creates the potential for increased sales and provides a hedge against the uncertainty of the effects of population change on sales in any one store. A coordinated strategy of outlet expansion also provides scale economics for advertising, promotion, and distribution. Another approach for retail growth is portfolio diversification of the types of stores operated by the firm. Dayton-Hudson followed this strategy by establishing Target Discount Stores to meet the needs of consumers not frequenting department stores. Retail growth through either local network expansion, increase in market coverage, or portfolio diversification requires the location of new retail stores. 102 The retail locational problem can be subdivided into two steps, generally evaluating the potential available in a competitive market and to determine the exact location for the retail facility. The multi-store retailer must evaluate specific site alternatives, based on both the site's potential and its place in a multiple store network, in order to assure an optimum network expansion (La Londe 1961). The gravity model (Huff 1964; Nakanashi And Cooper 1974) and the analog approach (Applebaum 1968; Rogers And Green 1979) are methods for selecting specific site locations. Multi-location models (Achabal, Gorr, and Mahajan 1982; Ghosh and McLafferty 1982) have been developed for more complex network site selection processes. Most of today's models are an elaboration of Huff's retail gravitational model. DL Huff prepared his model in 1962 with the mathematical formulation of the model as follows: Pij = $1; / £1 £4- 11 F 11 where: P5 = the probability of a consumer at a given point of origin i traveling to a given shopping center j; g = the square footage of selling space devoted to the sale of a particular class of goods by shopping center j; Tr' = the travel time or distance or costs involved in getting from a consumer's travel base to shopping center j; A = a parameter which is to be empirically to reflect the cf feet of travel time or distance on various kinds of shopping trips (researchers estimate at around 2) Simply stated the model indicates that the probability of any shopper choosing a particularretail centeriscqual totheratioofthe utilityofthatcentertothesumofthe utilities of all potential competing centers in the system. Specifically the utility or attractiveness of a center is directly related to the size of the center and inversely related to the distance separating consumers from the center. Whereas the Reilly law was Wu: indivj 103 deterministic in nature Huff developed a multiple utility function with two variables, selling space and travel time. Furthermore the model includes the effects of competition on the behavior of the shopper. This model also allows for irregular trade area boundaries to be formed thereby more realistically identifying reality compared to the more theoretical central place theories. Craig, Ghosh and McLafferty (1984) indicate that Huff was the first to suggest that market areas were complex, continuous, and probabilistic rather than the non-overlapping geomeuical area ofcentral place theory. Pearce and Wee (1985) discuss the fact that the model assumes that similar socioeconomic characteristics of shoppers will patronize stores in the same manner--a deficiency of the model. Also all retail centers in the trade area are included in the model. The concept of ”store opportunity set" needs to consider only those centers directly competitive in the market for similar goods. Gautschi (1981) expressed this concern as well. The calibration procedure has usually involved dividing the market area into a number of zones based on residential characteristics and traffic patterns. Tests have been used on randomly selected households in various zones who are then surveyed and observed regarding their shopping habits. So direct consumer behavior is used to reveal the pattern of preferences for alternative shopping opportunities. Various factors such as store image attractiveness and other traffic variables identified later in Table 16 have been used to test the reliability of the model and to enhance its forecasting ability. Achabal, Gorr, and Mahajan (1982) proposed the Multiloc model to assist multi-unit companies to find an optimum network of stores within a market area using a random search procedure combined with an interchange heuristic (using the Tomquist method) to identify optimal sets of locations. The model is not intended to select optimum individual locations only; rather it attempts to locate an optimum network of locations in order to saturate the market most prudently for overall market penetration and profit regardless of individual store performance. Whereas the MCI gravity model (Nakanashi and Cooper 104 1974) is used well for locating one store in a market given a level of existing competition, Multiloc is for maximum coverage of the market area with multiple units. Store growth is observed to occur in an S-shaped relationship between store share and market share within a region meaning that early market entry in full force is required for maximum market penetration-«but the question is where? Achabal et. al. surveyed 16 retail organizations regarding 26 of their major markets. Each firm evaluated 125 sites for an average 24 locations with the desire to open simultaneously for economies of advertising distribution and management. The researchers found that the retailers typically looked for good sites but did not attempt to develop a strategy for overall market penetration. The model is suggested for small multiple free- standing units such as convcnicncc stores, financial institutions, service stations, and supermarkets with limited use for large scale retailers. The model is a fairly detailed mathematical formula as an expanded Huff concept with an attempt through a heuristic approach to find the best overall system fit of store locations for complete market saturation. A given desired number of stores is provided in advance based on capital expenditure considerations of the firm with a heuristic analysis of various random store subsets within a market area. Then for each combination of store location subsets, total estimated consumer patronage sales and profit are estimated. Eventually a rank ordering occurs on the basis of store location combinations of store sales and profit. Although the initial analysis uses a set number of stores with real alternative sites, Multiloc also can be used for a determination of the total optimum number of stores by assisting in the determination of the point at which marginal revenue equals marginal cost. The Multiloc system attempts to assist in more easily focusing on the simultaneous evaluation of factors of size, image, location, and profitability of a group of potential stores. The authors believe that the model in concert with the firm's analyst, who has substantial applied experience, can assist in the limiting process of site evaluation and focus on those sites of highest priority for maximum penetration. 105 Goodchild (1984) proposed the ILACS model for evaluating site selection in regional areas as well as in metro locations. Whereas Multiloc is used for specific alternative site evaluations within a market, ILACS is used to evaluate alternative metropolitan areas and prepare a general strategy for the city as a whole primarily for low order functions. Conceptually, the model can be used to assist a market strategy of competitor avoidance or it can be used for direct competition. Goodchild indicates that for higher order goods where substantial overiap between service areas is likely it is necessary to calibrate and apply models of spatial consumer behavior in order to be able to predict market areas and total consumer patronage at the site. He believes that retail analysis, therefore, cannot easily assist in the higher goods analysis but can be very beneficial in the low order convenience goods where spatial trends tend to be nearest place with elasticity of demand based on distance. Goodchild describes the ILACS approach by using census data and geographic information system mapping procedures. The key factor is speed of travel versus distance with an evaluation of barriers of access. Again the Tomquist or the alternating heuristic algorithmic approach is used for optimum location solutions with recognition that identified locations are generic in nature and the necessity for the firm to identify real developable sites in the general vicinity. Whereas Multiloc looks at specific sites, the ILACS approach can be used in either a large region to pinpoint specific regional targets (c. g. localities within a state) or in a metro area providing general target locations. The attempt is to identify a general target area for the store analyst to then focus on specific sites. But again, this approach is deemed appropriate for nearest neighbor, low-order goods only. The various models either focus on specific location allocation strategies or they are for consumer patronage and market share evaluations. The Huff model serves as the elegant basis in all of these alternative methods for determining effective retail location strategy. Interestingly, the complexities of hi ghcr order goods market areas in metropolitan environs creates difficulties for determining effective models (compared to the relative ease 106 of the analog or judgment of the analyst by looking at supply/demand factors for the individual location such as a regional center) and they appear to be most useful for nearest neighbor lower-order goods strategies. Yet, even lower order goods retailing is significantly affected by multipurpose shopping thereby creating additional difficulties for model formulation. The use of linear multiple regression for modeling purposes is well-documented in both the applied and academic literature, although usually as support for the analyst to make experienced judgments. Wilson (1984) also supports the use of principal components analysis and other techniques. Ross Davies (1973) used both factor and cluster analysis to develop a classification for seventy-two chain stores that marketed durable goods in Britain. Obviously, multi-collinearity can be a problem and a thorough discussion of the statistical assumptions of the linear regression model is provided in Poole and O'Farrel (1971) and Ghosh and McLafferty (1987). They state that regression analysis can provide valuable insights into the determinants of retail performance in a variety of contexts at either the micro or macro level, although most have looked at specific site locations for comparative analysis. W While the above detailed locational models are helpful if a great deal of information is known about an area, shoppers trends, and store performance, Ingene (1984) indicates that considerably less academic attention has been directed to looking at the macro structural determinants of market potential for communities to enter prior to site selection. Structural determinants encompass those factors that are beyond the control of individual retail firms such as socioeconomic and demographic characteristics of households and the collective macro-marketing mix of retailers presently in the market, which Ingene determined through factor analysis including: up But {011m the d11 regard) 107 1. Assortment defined generally by square footage per store. 2. Service quality determined by average annual wage rate. 3. Service quantity based on employees per square foot. 4. Store density based on stores per square mile. 5. Atmospherics determined on stores less than five years of age. 6. % Mom-and-Pop Management based on percentage of stores with no paid employees. Ingene used secondary data in evaluating eight lines of trade including department and general merchandise stores of which discounters are a part. Findings varied significantly by line of trade. Ingene and Lusch (1980) demonstrated that basic demand variables like population and income account for relatively little of the cross-section variation in department store sales among SMSAs. But they do conclude that "department store expenditures per household can be strongly influenced by the marketing actions of department store managers." Ingene (1984) also finds that "household characteristics alone do not seem to explain a preponderance of the cross-sectional variation in per household expenditures." Ingene (1984) uses the following socio-economic variables, mostly noted in prior literature including income, age, age head of household, percent male, median age, household size, population, percentage white, and mobility defined as auto ownership. He factor analyzes the data and finds seven factors with three life-cycle, two income-related, and a racial and mobility factor. He then regresses those factors on per household retail expenditures with an Adjusted R2 = 0.3077 versus 0.47822 for the 31 original variables. But the Adjusted R2 for department stores on the factor regression was only 0.2237. Department store expenditures per household were significantly affected by only the following: Post-nest (-); f ull-nest (+); low income (-); and middle income (+). Part of the difficulty with a metropolitan analysis may be the nature of aggregation, especially regarding the racial data not being significant. 108 Ingene then adds the expanded model with the macro-marketing mix factors and performed regression on department store sales with an Adjusted R2 = 0.2981, only a slight increase in explanation with the following significant variables: post nest (-); full nest (+); low income (-); middle income (+); non-white (-); mobility -auto (+); assortment (+); service quality (+); store density (+); and atmospherics (+). In fact, the macro-marketing factors affected the explanatory power for department stores the least of eight lines of retailing. Ingene indicates that some of the variation in department store expenditures may be due to differences in the distribution of traditional, national chain, and discount stores. Undoubtedly, these are important considerations especially since his 1984 study used 1977 generally available date-when discounting was starting to expand tremendously. Ingene (1986) restates that "little empirical knowledge about retail structtu'e and its socioeconomic determinants has been attained” and that the lack of attention to the full breadth of micro-retail structure is due to "the absence of a theoretical or empirical explanation of why the statistical distribution of retail lines of trade differs across cities." Determining the distribution of stores across lines of trade requires an understanding of factors affecting store profitability, factors affecting the total number of stores in a line of trade, and the extent of competition among the lines of trade. Store profitability depends upon the gross margin being sufficient to cover overhead expense, which depends on the regional marketing concept that range of goods (stores) must be greater than the threshold value of the good (stores). The number of stores is an evolution of Christaller‘s (1966) central place theory of hierarchical distribution of goods, as expanded by Reilly (1931) in the retail gravity model. Ingene also describes the importance of urbanization economies regarding the costs of doing business. And very importantly regarding this analysis, retail structure is a function of past history and current realities, thereby suggesting that "the retail structure of high growth areas more accurately reflects current demand than does the structure of low-growth areas (Ingene 1986)." Little analysis has been performed regarding the effects of different lines of trade being in 109 competition. This factor has become extremely important with the new shopping formats and major shift to scrambled merchandising. Ingene (1986) performed cluster analysis, which did show regionality, and factor analysis of similar variables as before. He attempts to describe the importance of centrality or urbanization through a ”number of stores' variable, which has been illustratively shown to be increasingly a questionable variable compared to square footage. In performing retail market studies, most professional retail analysts focus directly on actual square footage of the competition as well as the nature of that competition although it is very difficult to obtain nationally on an up-to-date basis. Ingene performs ordinary least-square regression of the 12 socioeconomic factors upon each of the 15 macro-economic factors. He found an Adjusted R3 = 0.215 for department stores, again showing the limited explanatory aspects of the model. Socioeconomic structure factors significantly related to department stores include: full nest (+), poverty (+). unemployed (-), growth (-), and density-urbanity (-). The use of factor analysis must be considered in the evaluation of these trends because "used-car dealers" appear in conjunction with department stores and no reasons were offered. Also, data were for 1977 prior to major discount store growth. The signs do not really make sense for several variables. Poverty is a positive sign whereas unemployed is negative. Growth is a negative factor as well as density. Some of lngene's explanation for variation among all the retail lines may depend on excluded variables such as '( l) socio—economic factors such as homeownership, education, occupation and ethnicity, (2) various macromarketing mix factors such as store size, the capital-to-labor ratio, and advertising, and (3) environmental factors such as mass transit, legal restrictions on operating hours (blue laws), and the degree of state control exercised in the alcoholic beverage area.” Ingene and Brown (1987) looked at the strucuue of retail gasoline generally following the prior model approaches with 27 demographic and 14 environmental variables 1 10 to evaluate several dependent variables including number of stations per household and dollar volume per station. Environmental variables included size and density as indicators hypothesized to affect production costs (e. g. land and labor). They factor analyzed the data for 126 SMSAs extracting six components including full nest, seniors, poor, urban (size and density), manufacturing belt (percent of labor), and density. They note that their analysis does not include various important costs such as land rents, property taxes, leasing arrangements, utility fees, etc. Governmental regulations and taxes at the state and local levels have been shown to have an impact on retail structure, although not well understood (Ingene and Brown 1987). Ingene (1991) analyzes the first investigation of the impact of a set of structural determinants on the cross-sectional variation in the level of expenditures per household for ten spatial consumer services such as laundry, beauty and barber, auto repair, etc. It draws upon the regional science and economics literature pertaining to labor quality and it looks at the legal form of organization on expenditures per household, given the substantial nature of unincorporated businesses in this sector. He performs a similar principal components analysis and notes that the larger the household, the lower is discretionary income for services. Of key importance to this investigation is the statement that the more suburbanized an SMSA is, the lower are land rents. Furthermore, "establishments will be physically larger and will have more employees to operate.” He also suggests that population density will make land rents higher and travel more difficult thereby discouraging large stores due to land assembly and public transit comparison shopping difficulties. And low growth SMSAs will limit the size of establishments. Ingene finds that environmental factors such as population density and degree of suburbanization have been shown to be more important than previously thought. In addition to market potential, other factors need to be considered. Obviously, the economic base and future growth potential of the community need to be considered. The accessibility of the market area to the distribution system used by the firm is vital. Also, l 11 key factors include: availability and cost of media for advertising, availability and cost of labor, responsiveness of local government to new business, availability of credit, and the access to credit-investigation and collection services. The regulatory and legal environment, zoning regulations, and local tax laws must also be taken into consideration in making marketing decisions (Ghosh and McLafferty 1987). Determining the attractiveness of sub-areas within a metropolitan area for store expansion are based on the population characteristics and the competitive environment. Ghosh and McLafferty (1987) describe a portfolio matrix to determine the attractiveness of an area based on economic potential versus competitive position. Economic potential is a composite index based on such factors as population size, growth rate, income and spending patterns, and the degree of match between the existing characteristics and those of the firm's target market. Competitive position is from the competitive inventory. The two axes are demarcated in three regions representing high, medium, and low opportunities as found in Table 15. Table 15 Attractivene. of Subareas W High Medium Low W “Isl + + 0 M Medium + + Positive Attractive“ 0 neutral; - negative Source: Ghosh. Avijit and Sara I. McLafferty. locaflou 317023th and Service Fir-Is. Lexington: D.C. Heath and Company. 1987. The attractiveness matrix is a useful tool for screening different subareas and eliminating those unsuitable for new outlets. (For a general discussion of portfolio matrices see Wind and Mahajan 1981 with applications to site selection in Mahajan, Sharma, and Srinivas 1985.) 112 SelectedRe ' '0 Variables The following variables in Table 16 are from a review of over fifty recent journal articles focusing on retailing from both a macro and micro perspective. They are categorized by various typologies. While virtually all articles had demographics and time/travel factors, those authors cited herein are those who especially focused on these particular variables in their evaluations. These variables reflect the broad array of aspects pertaining to retail locations and shopper patronage. Summm The theoretical foundation for retail location has tended to focus on either site-specific analysis or on the metropolitan perspective of establishing an optimum portfolio of stores. Although many studies provided retail location variables, they are primarily oriented for evaluation of specific retail sites. Central place, gravity, and multi-locational models have been developed as locational models to be used when substantial data exists. However, Ingene (1984) has indicated in several analyses that considerable less academic attention has been directed to looking at the macro structural determinants of market potential for communities to enter prior to site selection. These previous analyses, including the use of multiple regression, factor, and cluster techniques were not successfully descriptive regarding metropolitan department store (including discount store) retail sales. No models were found that distinguished between central city and suburban retailing-most were metropolitan based from a national perspective. Furthermore, no analyses were found focusing specifically on discount stores. Also, studies typically focused on sales per capita or on a per store basis, ignoring the importance of square footage to provide a better comprehension of the scale of discount store space in a community. Thus, a model is formulated in the next chapter focusing on both discount stores and the differences between central cities and their related suburbs. This model formulation is intended to test whether central cities that are relatively comparable to their suburbs have attracted f ull-line discount store space commensurate with gross income potential. Table 16 Selected Retail Location Variables Patronage Choice 113 Motivational and experiential aspects of shopping Maximamutilitythroughrniaimamreaoarceexpenditaresand Time savings Browsing and getting ideas Mani-pupae Mains Recreational/leisurepnrposes Coatofgoodsfin-‘orccoatsandstoragecoats Prodaetriskforeeonomicandsocialrishofgoods Merchanttrustworthinemandservioe Qaalityofloeation 3““? Travel I Tramportation Driving time and distance Travel con Travel cedar! Number of aatoa per household Parking availability and ease Convenience to work Transportation infrastructure Demographic Tradeareademographicsaadgrowthuends $singlefamilyhoasing Rnceage.andineome Typeofeanomersandlifestyle Socislclauofcastomera-realandperceived w (Anglia eul. r991) (Anglia an. 1991; Ghmll and McLafferty 1986) (Holmn and Wilson 1984: May 1%9) (Anglia etal. 1991; Bloch eul. 1991) (Anglia eta]. 1991; Ghosh and Ingene 1991: Ghosh 1986; Ghosh and Melafferty 1986; Holman and Wilson 1984: Mulligan 1%7) (Anglia eul. 1991; Bloehetal. 1991) (Ghosh and Ingene 1990) (Kora-01m! 1982) (Gautschi 1981; May 1989) (Achabal etal. 1982; Anglia et.al. 1991: Black 1%; Goodchild 1984: Granboisl981; Howell and Rogers 1981: Kohsaka 1989 (Anglia etal. 1991: Bellenger eul. 1977; Bloch eul.1991; Gautschi 1981) (An ' etal. 1991; Black 1%4; Barns andGentry 1977; Fein et.al. 1991; Ghosh and Ingene 1991: Howell and Rogers 1981; Kern etal. 1984; Pearce and Wee 1985; Zikmand 1977) (Black 1984; Gautschi 1%1; Ghosh and Ingene 1990) (Oamaehi 1981: Mahajan et.al. 1985: Snith 1%5) (Holmananderson 1W4: Ingene 1984. 1%6. and 1991; Kern et..al. 1984) (Mfineul. 1991: BarnsaadGeatry 1977; Howell andRogea 1% ) (Bellenger eta]. 1977) (Anglia eta]. 1991: Clements 1978; Gautschi 1%1; Ghosh and McLafferty 1991) (Anglia et.al.l991) (Doherty 1991; Holman and Wilson 1984:1agene 1M 1%6. and 1991;1ngeneandLaaeh 1m; Kohsaka1989; Kosgaonkar 1982; Mahajan et.al. 1985) (Mahajan eta]. 1985) (Hall 19$; Ingene 1984. 1986. 1991;1ngene and Lusch 1980. Lloyd andJeanings 1978) (Baruand Gentry 1977; Doherty 1991; May 1989) (Dawson etal. 1990) 114 Table 16 (coat'd) Physical(bytypeofcentersuchaspowerhi h fashion/specialty mini mega regional off-prr’geloatlet hypcrmarketlaupemore areasldistricts centers nulls etc.) Broad array Vacancy rates Surrounding neighborhood Quality of center Facility age Centeraiaeandmsss Buildings and landscaping Image Shopping area image Shopper image and reputation Affectorpositivestatemood Formlityofdre- I"| Financial imtitutional investment crategies Stores and developer consolidation Advertising program and media Governmental Zoning Politics and government seeking revenue Atmosphere Clean visually exciting attractivepersoanel safety atmosphere Visitornirnuliandexcitement Affiliation with others and group gathering; planned events Amortment Store quality variety of merchandise latest fashions Variety and value Storerrs'x Econonu'cs Price levels sales Land rent (Anglia Gentry and Stohmsn 1991) (Mahajan seat 1985) Mill M 1985) (Bellenger eta]. 1977; Burns and Gentry 1977) (Black 1984: Bloch etal. 1977) (Achabal et.al. 1982 An ’ et.al. 1991; Black 1%; Fein et.al. 1991; Hall was; et.al. 1984; Pearce and Wee 1985) (Burns and Gentry 1977; Gautschi 1%1; Ker-a et.al. 1984; Kohsaka 1989: Wee 1986) (Anan et.al. 1991) (Anglia etal. 1991; Burns and Gentry 1977; Doherty 1991; Ogledge and 'I’immermarn 1990; Granbc'l 1981; Lsagrehr 1 l ) (Golden and Zrmrner' 1986) (Gautschi 1981) (Brown 1989; Ghosh and McLafferty 1991) (Ghosh and McLafferty 1991: Ingene 1%6; May and McNair 1977; Nowakahtar and Widdows 1983 ) (Black 1%; Burns and 1977; Doherty 1991; Howell and Rogers 1981; Peareeand cc 1985: Zihnnnd 1977) (Brown 1989:1(ohsaka 1989; Mahajsaetal. 1%5) (Brown 1989: Ghosh and McLafferty 1991) (Anglia etal 1991; Black 1984; Burns and 1977; Gautschi 1%1; Howell and Rogers 1981;1ngene 1%4and 1991:1rsngrehr 1991: Smith 1985; Wee 1986) (Bloch etal. 1991) (Blocheul. 1991; Smith 1%5) (Anglia etal. 1991; Black 1984: Bloch et.al. 1991; Feinberg eta]. 1991; Gautschi 1981; Howell and Rogers 1981;1ngeae 1984; Wee 1986) (Anglia eta]. 1991; Burns and Gentry 1977: May 1989) (Anglia eta]. 1991; Anderson 1%5; Bellenger et.al. 1977; Brown 1989; Ingene 1984) (Anglia etal. 1991; Burns and Gentry 1977: Gautschi 1%1) (giggle and Brown 1987;1ngene 1986and 1991;1(ohsaka 115 Table 16 (cont'd) Amenities and Conveniences Shopping layout special exhibis (Achabal et.al. 1982; Anglia et.al. 1991) Not crowded (Anglia out 1991; Gautschi 1%1) Euytowalkaroundgoodhours (AnglineLal. 1991;8ellengeret.al. 1977:BurnaaadGentry 1977: Gautschi 1981; Wee 1W6) Credit card service (Achabal et.al. 1982) Lowentrycot (Bloch etal. 1991; Ingene 1991) Food and refreshments restroom (Anglia et.al. 1991; Wee 1%6) Related services such as banks theaters restaurant (Bellenger eta]. 1977) CHAPTER 5 PROPOSED MODEL, HYPOI'HESES, AND METHODOLOGY erview The approach to the geographical research incorporated herein somewhat follows the concepts of positivism, with its associated scientific method (Harvey 1969). Generally, this approach includes a logical set of procedures from theory, model formulation, hypotheses, data collection, hypothesis testing, and evaluation. However, given the immaturity of the social and environmental sciences, few axioms really exist upon which to base theoretical development. Geographers tend to make assumptions and then test their validity before deductions can be made (Johnston 1991). The general theoretical foundation, albeit limited in this particular realm of retailing, was established in the literature review with this chapter establishing the model and hypotheses. The Demndent Variable A reasonable way to look at how much discount department store space a central city should have, or would like to have, is to consider the city within its demographic context in the metropolitan area. At a minimum, a city would prefer to have at least the same percentage of space as its percentage of metropolitan population. Recognizing that income is an important variable affecting demand, total population might be adjusted by income such that space in a city should be proportional to the city's gross income (Population "‘ Effective Buying Income) as a percentage of the metropolitan gross income. At a minimum, the city would like to capture the economic demand theoretically created by its own residents. 116 1 17 Thus, CAPTRLQ is the dependent variable reflecting: Cig Sguare Feet of Discount Dement Store Sm ; C'g Gross lrmme Metropolitan Square Feet Metropolitan Gross Income Simply stated, if the city has 50% of the metro area's gross income, then the city should have 50% of the discount store space to have 100% capture. If the city only has 10% of the metro area space in this scenario, then the city would have 20% capture; or if the city has 60% of the metro space, then the city has a theoretical capture of 120%. Some cities may actually be doing better than 100% capture due to attraction of suburban and other non-resident shoppers, although discount store shopping tends to be more residentially-based than say department stores, which might attract tourists and a broader regional trade. Those cities not even capturing their own percentage of metropolitan gross income would really not be performing well. Certainly, lower income households in central cities would not have the same personal consumption expenditures for discount store goods as the median metropolitan or suburban household, but this will be reflected in one of the independent variables. This dependent variable provides the ability to evaluate metropolitan portfolios for discounters by looking at the city versus suburban competitive supply and demand factors. If per capita space in central cities had been chosen as the dependent variable, then the analysis would have been altered to strictly a comparison of central cities, regardless of their relationship to their suburbs. Although central city planners may be interested in how they compare to other central cities, they are primarily interested in how their cities perform within their own metropolitan perspective (i.e. CAPTRLQ). 1 18 W1 The theoretical model argues that the ”Capture Location Quotient“ by a central city is a function of variables affecting the demand for and supply of discount department store space, such that: CAPI'RLQ = f (LA, 10) where: CAPTRLQ = the percent of space in the city related to its gross income compared to the metropolitan area's space and gm income (generally described as "Capture location Quotient”) LA = a vector of three variables pertaining to the availability and cost of land assembly 10 = a vector of four socio-economic and crime variables comparing the city to its suburbs as a basis of estimating ”investment opportunity" in the city The model is an evolution of prior work done primarily on a metropolitan level, but no distinct body of theory has been observed at the central city scale, especially regarding actual square footage at this level of industrial classificationuthe discount department store. This model considers the importance of the metropolitan context for the central city as a key factor in discount store location and portfolio strategy. The seven independent variables, select descriptive statistics, and their sources are identified in Table 17. Table 17 Independent Variables (N27) 119 symbol Brief Description Mininm Maximum Median Mean Source Land Amembly C%80-90H mfifinfiw 1.8 34.1 10.2 13.1 Cells CDENSl'I'Y mm Benjy/Square 0.7 23.7 4.6 5.9 Census SALES%TAX Sales ¥$m as a ”b of 0.0 82.6 30.3 28.7 CGF Cetus Investment Opportunity HHEBILQ City Household EBI: Suburb 0.49 1.06 0.72 0.73 8’ ¢liBl.1(LQ City % Black:Suburb 1.09 43.57 3% 53 Census %HSH.Q City % Hispanic18uburb 0.71 5.75 1.59 2.06 Cemas VIOLLQ City Violent Crime Index 1.3 10.2 3.7 4.1 FBI at“ to Suburban Violent Sauces: Censua=U.S. Celsius ofPopulation and Housing. 811‘ l: SBP=Sa|es& Marketing Managemnt . Sterne/Buying Power. 24 August 1992; GOP Census=U.S. Bureau of the Cenatt, City Gourmet Finances, series GF. No. 4. annual; FBI=U.S. Department of Justice. Federal Bureau of Investigation. cm is the United States. 1992: (lam Crime Report,30ctober1993. flmtheses The multiple linear regression. model is used in this analysis expressed as: Yi=l30+51X1r+BzX2r+---+Bpxpi+€1 The B terms are unknown parameters and the 81 terms are independent random variables tint are normally distributed with mean 0 and constant variance 02. The model assumes that there is a normal distribution of the dependent variable for every combination of the values of the independent variables in the model. The following hypotheses and rationale for each variable are described. Hozfil-Bz-...-B‘-0 versus Ha :atleastoniju-O The alternative hypothesis is : CAPI'RLQ = CONSTANT + C%80—90H - CDENSITY + SALES%TAX + HHEBILQ - %BLKLQ - %HSPLQ - VIOLLQ 120 Briefly, the Capture Location Quotient is dependent upon the following independent variables. 1. Positively related to the amount of new housing construction over the last decade. This variable reflects new growth and the likely availability of land for commercial purposes simultaneously with new housing development. 2. Negatively related to density, which is an indicator of both limited land availability and high cost of acquisition within the city limits. 3. Positively related to sales tax as a source of local revenue for the city since cities more dependent on sales tax will be apt to create varying degrees of public-private partnerships that assist in land assembly or availability. 4. Positively related to cities that are closer in household income to their suburbs. Since the suburbs have been the premier location for shopping center and discount store development, the hypothesis is that this development will also be located in central cities more closely resembling the income characteristics of their suburbs. 5. Negatively related to the percentage of Blacks and Hispanics who are in the cities compared to their suburbs due either to discrimination, income disparities, or unwillingness of developers/discounters to penetrate heterogeneous markets. 6. Negatively related to the amount of violent crime compared to their suburban counterpart. Thus, those cities that have more violent crime than their suburbs will attract less discount store space. These variables and related hypotheses are described more fully below. Land Assemny ngables LA, or land assembly, consists of three independent variables reflecting the ease and cost of land assembly. Land assembly at a reasonable cost has been identified throughout this research as a major problem for central city development. Without performing individual city market studies, site availability at a reasonable cost cannot be specifically 121 identified. But these three variables are shown to be good indicators of land availability and the related cost parameters. C%80-90H is the percentage of housing structures built in the central city between 1980 and 1990. The fifty central cities range from 1.8% to 34.1% of their total housing structures having been built in the eighties with a median of 10.2%--a substantial variation. Discount department store sizes have increased considerably during the last decade with the standard full-line store prototype (increasing from 60,000 to 100,000 square feet) and with the shift to the development of supercenters in the 150,000 to 200,000 square foot range. Furthermore, whereas they previously were free-standing units or in small neighborhood convenience centers, many are now in large power shopping centers of between 400,000 to 600,000 square feet. Other new stores tend to be part of large-scale shopping agglomerations with nearby competitors or adjacent to regional malls. Some are even becoming part of regional malls as anchors. Thus, the large-scale land requirements of modem shopping facilities and discount department stores illustrate that land availability has become even more important today for development, with sites of between 20 and 50 acres typically required. Hall, Knapp, and Winsten (1961) indicated in areas of growing population, retail stores experience an "accumulation of profits,“ capital is attracted from other areas, and an entrepreneurial attitude of excitement exists that leads to less traditionalism. Because of the greater availability of capital in high growth areas, more stores are new in modern formats and older stores are renovated/expanded. Profit is created significantly by atrnospherics, both exterior and internal, and investment tends to be in positive growth areas (Ingene and Lusch 1980). Although Ingene and Lusch selected rate of population growth as a proxy for atmospheries, housing development provides a better identifier for both the atrnospheries and land availability since the assumption is that land for shopping center development would be available simultaneously with new housing growth. 122 This variable is also a good indicator of the life cycle process with relative homogeneity of new development areas from a residential neighborhood (i.e. trade area) and commercial investment standpoint. Also, the variable acts as an indicator of f all-nest households since approximately 65% of the new units in the eighties were single-family. Thus, (DCAPI‘RIQ,C%80-9OH >0)- In addition to land availability, the cost of land is important and has been identified in various studies as being a density factor. Many cities have been allowed to annex through the years or have grown in p0pulation recently compared to older upper midwestern/northeastern communities. CDENSITY, which is population per square mile in the city, serves as a proxy for land cost under the assumption that less dense land means that raw land is available or that existing low density areas could be redeveloped more economically compared to dense cities. The fifty central cities range from 731 people per square mile to 23,698 with a median of 4,641. This variable represents not only cost but likely availability of land. It is impossible to know specific sites and land availability near acsessible and good trade areas for each central city market without a detailed market study. Whereas density suggests that discounters would like to be near population agglomerations, limited land availability has ’ precluded development from occurring. Only very recently have discounters been willing to enter multi-level structures in order to deal with the unavailable land, although land rent is substantially higher than typical suburban sites. Both CDENSITY and C%80-90H are good indicators for land availability and cost. Rather than use actual land area in the city limits as a variable, CDENSITY provides a better proxy for land cost by illustrating the degree of urbanization within the city limits. Therefore, (ngaqsm d1). Central cities have various taxing capacities including the opportunity to attract sales tax receipts for retail goods sold in the community, regardless of where the shopper resides. Some studies have shown that general merchandise stores in a city (especially 123 with strong downtown retailing) can generate substantial general merchandise sales «and related sales taxes--frorn tourists or suburbanites who do not reside in the city. The city of St. Louis estimates that 30% of its general merchandise sales tax revenue is from non- residents. Just the opposite is also possible if the city does not have adequate shopping facilities with central city residents going to the suburbs to shop. Determining tax rates and tax opportunities in cities (including sharing state tax collections) are beyond the scope of this study, but an indicator of the importance of sales taxes to the community can be found, although it is somewhat correlated with the dependent variable. The SALES%TAX represents the sales tax income as a percent of the actual local tax revenue generation in 1991. Although all cities receive some property tax, not all cities have the legislative authority at either the state or local level to levy sales taxes. Thus, the hypothesis is that cities with higher sales tax revenues will encourage modern discount store development. Craig, Ghosh, and McLafferty (1984) cite this as an important factor for suburban governments encouraging shopping center development. The fifty central cities range from no sales tax revenue to 82.6% of the local tax-generating revenue base with a median of 30.3%. This variable represents the political factor in attracting development with its employment and tax creation impacts. Cities may encourage development through zoning or an array of public/private partnership incentives such as land writedown, loans and grants, and property tax abatements. Many cities have attempted to find alternative sources of revenue to the prOperty tax that can also attract non-residents such as the sales tax. It shouldbenotedtlratonanintraurban basis,thelocal salestaxmaybeahindranceto encomaging development if the city's tax rate is out of line with competing metropolitan sites. But the hypothesis is that cities with sales tax revenue as an important percentage of their gross revenue will try to encourage development. Therefore, (WWW-[Ax >0). 124 nves t rtuni Varia es Central cities do not operate in isolation and are dependent partly upon their relationship to the suburban environment of their metropolitan areas. Given the perceived/real bias towards central cities by investors, I hypothesize that the closer the central cities' characteristics are to their suburbs, the higher the CAPTRLQ will be. The hypothesis is that those discounters and shopping center developers active in a metropolitan market will be more favorably disposed towards investing in a central city that resembles its suburbs. Four independent variables have been selected as indicative of ”investment opportunity.’I Income, race, and crime are pervasive factors affecting city-suburban relationships. Although this research describes the evolution of micro-marketing and market segmentation for discount store operations, relatively homogeneous areas have been selected as site locations until recently. Given the market saturation in selected suburban areas, it is hypothesized that portfolio expansion of stores into central cities would follow a spatial strategy of penetrating cities most like their suburban portfolios. Trade area characteristics for central city locations comparable to their suburbs represent considerations of the life cycle concept for investment. As described in the literature review, income, race, and crime have significant effects on neighborhood and commercial life cycles-hence, investment opportunity. The following four independent variables are the ratio of the central city versus suburban data. These fifty cities vary considerably within their metropolitan context. HHEBILQ represents the city's median household effective buying income (EBI) as a ratio with the suburban household EBI. Compiled by Sales Marketing & Management, EBI is defined as personal income less personal tax and non-tax payments-a number often referred to as ”disposable" or "after- tax" income. Median household data are deemed to be a more reliable indicator of an area's income level and less likely to be skewed than mean income. With higher household 125 income, a portion of the increase will be spent in discount stores creating additional demand. The city's CAPTRLQ will be positively related to the ratio. For example, if the city's EBI is $25,000 and the suburban EBI is $50,000, then the ratio will be 0.5. If the city were $50,000 instead, then the ratio would be 1.0 and the city would have a better percentage capture. These cities range from 0.49 to 1.06 in their cityzsuburb relationship with a median of 0.72. Only two cities have a higher median income than their suburbs. Thus. (bCAPl'RLQHHEBILQ >0)- %BLKLQ represents the amount of African-American integration (at least in the aggregate) in the community with the hypothesis that the closer the city resembles its suburbs racially, the higher the CAPTRLQ will be. Similarly, %HSPLQ represents the Hispanic ratio. The %BLKLQ for the fifty cities ranges from 1.09 to 43.57 with a median of 3.86 and the %HSPLQ ranges from 0.71 to 5.75 with a median of 1.59. Obviously, these communities vary substantially in their metropolitan racial diversity. Therefore, (bCAPFRIQ.%BLJQQ <0) and (W.%HSH_Q <0). Percent non-white was considered, but the distinction between the two are deemed significant with %BLKLQ hypothesized to be more negative. The negative relationship is predicted due to a variety of factors including racial discrimination by shoppers or developers/tenants, lower income effects, and inability or lack of interest on the part of discounters for marketing to a heterogeneous segment of the market (either due to a capital or operating investment risk or due to discrimination). These variables are also an indicator of life cycle patterns of neighborhoods based on invasion-succession theory. VIOLLQ is the ratio of the city violent crimes index to the suburban area. Crime has not been included previously in metropolitan-oriented retail studies given the level of aggregation, but it is an important factor from a psychological and real perspective to consumers and investors in the context of city versus suburb. Of course, crime figures can 126 vary considerably by city neighborhood for individual store sites but the overall crime picture would still be negative. The variable represents violent crime per 100,000 population for 1991. Although total crime is important, many analysts believe that the fear of violent crime is the more important consideration because it reflects on personal safety. Violent crime includes offenses of murder, forcible rape, robber, and aggravated assault. Even though the degree of violence in each city may also be important, the hypothesis is that investment will be affected more by the city-suburban ratio than the actual crime index for the city itself. These cities range from 1.3 to 10.2 in VIOLQ with a median of 3.7. Therefore. (bCAPrRroyrouo <0). Methodological Overview Johnston (1991) indicates that much of human geographical research is not based on a properly constituted sample, but on a given population thereby creating limited ability to undertake inductive statistical analysis. Therefore, significance testing for inferences to populations is hampered by the nature of the research; but geographical research can use such techniques for descriptions of relationships within the context of causal hypotheses. When a population is being analyzed in a regression equation, the various coefficients and correlations are exact descriptions of the trends in the data set within the parameters of the linear model. Johnston suggests, however, that the F-ratio may be used by researchers not using sample data as a check on validity. Furthermore, the t values may be used on the argument that they indicate the relative strength of the relationship in a regression. And the parameters of regression equations are valid for descriptive purposes when they refer to data that meet the requirements of the general linear model. Some critics have attacked use of certain techniques . . . because these are not used to test hypotheses, but only to answer very vague questions about relationships and differences. The answer offered to this attack is that in a state of immaturity, use of these methods in a sensible way allows geographers to crystallise their ideas and so to move slowly towards the adoption of the formal methodology. At present our notions of causes and relationships are vague and data are being processed to improve on this vagueness, to provide useful foundations for scientific analysis (Johnston 1991). 127 i 'onof tStud reas taCol ' Ingene and Lusch (1980) suggest that data analysis at the SMSA level is appropriate for two reasons. First, SMSAs are defined to be economic integrated areas as labor markets and where few retail sales are made to residents of nearby rural areas that are not a part of the SMSA. Furthermore, newspaper and television advertising and marketing influence tends to diminish beyond the SMSA borders, thereby establishing the geographical range for store competition. SMSAs became Metropolitan Statistical Areas (MSA) in 1983. MSAs generally exist for all cities over 50,000 and are defined in terms of counties, based on their urban population, p0pulation density, and share of workers who commute to the central county(ies) of the metro area (Hartshorn 1992). I have disaggregated data into “central city” versus ”suburban" for the top fifty Metropolitan Statistical Areas (MSA) and Primary Metropolitan Statistical Areas (PMSA), which have central cities within their boundaries. Thus, consolidated metropolitan statistical areas (generally areas with two or more adjacent metropolitan areas sharing a common boundary) have been disaggregated into primary statistical area components, in order to allow for consistent market evaluation. Furthermore, central city data have been collected on only the major central city portion by which the MSA is known generally to the public. For example, St. Louis MO. has several communities that qualify as central cities in the MSA such as Florissant, Alton, Granite City, and East St. Louis; however, for this analysis, those secondary central cities are aggregated in the "suburban“ category since they tend to function more like suburban communities and they are not the focus of this analysis. Unless otherwise noted herein, the central city data are from the actual or shortened name of the PMSA cited in the tables. Selected differences include the following, based on the review of the data and desire to focus on generally acknowledged central city areas, although definition is somewhat arbitrary: 128 PMSA CENTRAL CITY *Los Angeles-ng Beach Los Angeles *Riverside-San Bemadino Riverside *Minneapolis-St. Paul Minneapolis and St. Paul *Anaheim-Santa Ana Anaheim *Tampa-St. Petersburg-Clearwater Tampa, St. Petersburg, and Clearwater *Miarni-l-Iialeah Miami *Kansas City Kansas City, MO. *Norfolk-Va. Beach-Newport News Norfolk *Fort Lauderdale-Hollywood- Fort Lauderdale Pompano Beach *Fort Worth-Arlington Fort Worth *Bergen-Passaic Paterson City *Charlotte-Gastonia—Rock Hill Charlotte *Salt Lake City-Ogden Salt Lake City *Greensboro-Winston Salem-High Point Greensboro and Winston-Salem *Dayton-Springfield Dayton Three PMSAs were not included because they are suburban in character and did not have central city data or had a “central city” of less than 50,000 population that is not generally known. They are Nassau-Suffolk (2.7 million ranked 12th); Middlesex- Somerset-Hunterdon (1.0 million ranked 46th); and Monmouth-Osean (1.0 million ranked 48th). Also, 44th ranked Hartford-New Britain-Middletown-Bristol (1.0) million was deleted because PMSA information is unavailable and only CMSA data exist. These four exclusions allowed Oklahoma City, Buffalo, Greensboro-Winston Salem-High Point, and Dayton-Springfield to be included in the top fifty. For final clarification, data are disagreggated into ”central city" and "suburban“ designations. But the suburban designation must be understood as being the remainder of the MSA or PMSA outside of the identified central city boundaries. Some portions of ”suburban” may be very much city-like in character and may even include secondary central cities within the MSA. My focus, however, is upon the key older central city known by the public versus the remainder of the MSA. Market Digggggtion Issues And Assumptions This analysis focuses on the public policy perspective of discount department store locations in central cities. Each metropolitan market may eventually need to be analyzed on 129 a micro-analysis basis regarding specific site locations. But this analysis attempt to identify broad-scale structural trends. The intent of focusing on PMSAs is to delimit a likely trade area for evaluation; however, disaggregation into sub-areas within the metro area provides some methodological problems due to "leakage" that can occur from cities to suburbs and vice versa. The disaggregation problem becomes more of an issue if smaller metro areas with smaller central cities had been included. Since most of these top fifty metro areas include large central cities (minimum is 140,900), the assumption is that sub- area demand should create the opportunity f or an adequate sub-area supply of discount store space. Therefore, disaggregation should not be problematic from a broad portfolio perspective. For example, conclusions could be affected if many discounters had a store one block from the cenual city boundaries in the suburbs. Even if this locational pattern existed, the analysis is still relevant from the perspective of the city's economic base being negatively affected although such locations might meet the shopping needs (demand) of the city resident. Given the 2,435 discount stores in the fifty metro areas, further research at the micro-trade area level might be beneficial at a later date. The assumption is that the locational pattern of these stores would not be biased regarding locations immediately adjacent to the city limits. Another important issue is the cross-sectional aspect of the analysis. As described herein, the discount store industry is in a dramatic growth and restructuring phase with numerous store closings, expansions, replacement, and new unit being developed annually. Several hundred unit are affected annually. For example, Kmart revised it 1990 plans to relocate 500 more stores and expand 270 by 1996. And the oldest unit being restructured may be in central cities and inner ring suburbs. So, this static analysis for 1992 needs to be considered in light of these changes as an indication of the existing portfolio of stores. Perhaps the analysis will show the store locations and sizes to anticipate likely restructuring effect, which is privately-held proprietary information. Also, the cross 130 sectional analysis evaluates an existing perspective in light of investment decisions made over the last forty years in the discount store business, especially the last twenty years of expansion. Trade areas, competition, store format, and other variables have changed. Yet, this analysis includes prior decisions from a static versus time series perspective. W Data sources are secondary published resources and are identified in the appropriate tables. They consist primarily of the following sources: * U.S. Department of Commerce, Bureau of the Census, 1990-1993 * U.S. Department of Justice, Bureau of Justice Statistics, 1992 * U.S. Department of Justice, Federal Bureau of Investigation, 1993 * Gale Research Inc., Market Share Reporter, 1992 " Lebhar-Friedman Inc., Directory of Discount Department Stores , 1992 “ Sales & Marketing Management, Survey of Buying Power, 1992 " Recent issues of trade publications including Chain Store Age Executive, Discount Merchandiser, Discount Store News, Monitor, and Shopping Center World. Much of the data require alteration in order to separate data into suburb and central cityausually by deleting the central city data from metro area data. No observations are missing in the data collection. Tables and transformations have been prepared on Macintosh with Microsoft Excel and Word, with statistical analysis in SPSS. The Directory of Discount Department Stores was provided by the company on disk format and includes every discount store in the nation in 1992 including such data as metro area, county, city, zip code, and size. As noted previously, the industry is in a major state of flux and errors undoubtedly occur in the compilation of over 9,000 discount stores nationwide. However, this source is considered by the industry to be the best publicly available. Unfortunately, some square footage estimates are missing for individual unit (less than 10% of total stores). The largest chains were contacted for confirmation of 131 missing square footage on select stores in the directory, although some chains considered the information proprietary. For the smaller chains (which tend to have similarly sized units and reflect a small percent of space in the area), I have assumed that any missing store data would be the average size of all stores for that chain in the individual metro market. For the larger chains, the store average nationwide has been used or an estimate was made based on judgment of the store type, if discernible (e.g. Wal-Mart's supercenters would equal others). Another difficulty with the directory is the ability to determine whether the zip codes are actually within the central city limits. Older city zip codes tend to follow city boundaries but some boundaries may extend slightly into the inner ring suburban area. Claritas was employed to prepare a 1993 zip code list by city with the percentage of population included in the city limits for each zip code. Discount stores in zip codes with 100% of the population in the central city were included automatieally. All other stores necessitated a telephone eall to each store to determine if the store was in the central city or the suburbs. While slight error might exist, all reasonable precautions to verify data were attempted and are deemed reliable. Data Analysis gd flmtheses jljesting Procedgrg As stated in the research pr0posal, data analysis procwded in the following manner: 1. An ordinary least-squares multiple regression analysis using the seven independent variables are regressed against the dependent variable, CAPTRLQ. Analysis is performed in SPSS for the Macintosh. Variables are inserted by forced entry in order of the largest positive or negative correlation with the dependent variable. 2.Thesignifreanoeoftheoverall modelthathasbeenfrttedtothedataistestedby Snedecor's F test examining the ratio of explained to unexplained variance. The F statistic is tested against [F, p = .95, df= 7, 42] = 2.24 132 3. The test for the signifieance of the partial regression coefficients follows the Student's t statistic with all T-tests one-tailed at the .05 signifieance level with 42 degrees of freedom. 4. The adjusted R2 is evaluated for the goodness of fit of the model. 5. Residuals are analyzed to determine any underlying pattern and to start the evaluation of possible assumption violations of the validity of the linear model. Mapping of residuals also assists in possible reformulation and modifieation of hypotheses, including recommendations for adding new independent variables or transformations to satisfy the assumptions of normality and linearity in future testing. Evaluation of residuals also assists in finding any areal association between a dependent variable and the set of independent variables. Residuals analysis provides a basis for selection of particular areas for future field investigation. Mapping also assists in determining whether spatial autocorrelation exists. Studentized and standardized residuals are evaluated in tabular format. Plots of the residuals against Y or X are prepared to evaluate normality by looking for a uniform band of equal variance for all values of the plotted variable. Deviant forms may assist in possible model reformulation. A histogram of studentized residuals also assists in determining normality. 6. A partial regression plot helps determine model inadequacies and violations of the underlying assumptions. 7. Outliers are examined for possible rejection or further evaluation, as determined by a review of the table of residuals and leverage values, which are the diagonal elements of the hat matrix. Leverage values exceeding (2p) I N, where p = the number of independent variables are examined for possible exclusion through plots of standardized change or the covariance ratio. 8. Each independent variable is tested in a seatterplot to determine if it has a linear relationship with Y. 133 9. The plot of observed Y against a selected X5 is evaluated for homoscedacity, the assumption of equal variances. 10. Multicollinearity is measured by the tolerance of the independent variables, which is the multiple correlation (squared) of Xj with all of the other independent variables in the regression, subtracted from 1. The tolerance value will vary from 0 (where Xj is a perfect linear combination of the other independent variables) to 1.0 (where it is uncorrelated with any of the other independent variables). The reciproeal of tolerance, variance inflation factor, will be reviewed for collinearity. 11. Upon completion of the test for the proposed model, forward, backward, and stepwise selection procedures are used for model reformulation and evaluation. Statistieal output includes multiple R data, analysis of variance table, statistics for variables in the equation, tolerance, select descriptive statistics, and residual measures. Data are provided in tabular form and in text description and evaluation. Summa_ry A linear model has been hypothesized incorporating seven independent variables representing structural aspects of investment opportunity and land assembly. Variables include new housing development, density, income, race, sales tax revenue, and violent crime characteristics. Discount store square footage in 1992 is used with the recognition that a cross-sectional analysis of a dynarrrieally changing discount store environment must be considered in drawing conclusions and recommendations. The industry is changing rapidly although the literature review suggests that most of the discount restructuring is still occurring in the suburbs. Data analysis techniques appropriate for multiple regression are identified prior to testing the model. CHAPTER 6 TEST EVALUATION AND DISCUSSION I est Besglg The resulting equation from the multiple regression analysis follows with T-values listed below the coefficients. CAPTRLQ = 22.2737 + 3.5390 (C%80-90H) -l.4339 (CDENSITY) - (0.512) (3.694) (-1.081) 0.1293 (SALES%TAX) + 10.2973 (HI-IEBILQ) + 1.1240 (%BLKLQ) - (-0.465) (0.187) (1.312) 3.1603 (%HSPLQ) + 1.1579 (VIOLLQ) (-0.710) (0.389) This equation has an F-Ratio of 7.25508 with Signif F = .0000, with an adjusted squared multiple R of 0.4719 and a standard error of 32.6462. The standard error is fairly large since it represents the CAPTRLQ that is assumed to be 100.0 if a city has its ”fair share.“ The range of CAPTRLQ for all fifty cities is 0.00 to 189.94 with a median of 63.85 and an average of 68.96. Table 18 provides the analysis of variance. Several key steps were performed to determine whether the assumptions for the use of the regression model were met. Scatterplots of the independent variables with the dependent variable show broad seatteration although C%&)—90H has a general linear tendency. SALES%TAX and %I-ISPLQ show the least linear patterns with the dependent variable. A correlation matrix of the variables did not appear to show excessive correlation with only two variables having coefficients over 0.6, C%80-90H with HHEBILQ (0.76) and C%80-90H with CAPTRLQ (0.70). However, multiple collinearity is indicated with 134 135 C%80—90H since it has the lowest tolerance level at 0.296 and the related VIF of 3.373; this compares with several variables having the highest tolerances at 0.63. Plots of the residuals against predicted Y, actual Y, and the independent variables did not show any nonrandom patterns. Neither the standardized nor studentized residuals were over (2.0) with only 16 of the 50 studentized residuals being over (1.0). However, ten of these sixteen were negative residuals. The histogram of the standardized residuals show proximity to the normal curve, although slightly skewed negatively; and the normal probability plots of both the standardized and studentized residuals as well as predicted values are relatively linear. Only one city, Milwaukee, has a high leverage of 0.68, which is over Belsey et. al.'s (1980) recommended threshold of 3K/N(i.e. 0.42). New York, Chicago, Cleveland, and Charlotte have leverage scores between 2K/N and 3K/N. The null hypothesis must be accepted for all independent variables other than C%80- 90H, for which the null is rejected and the alternative hypothesis is accepted. Given the significance of only C%80-90H, I performed a regression on that variable with the resulting equation having an F-Ratio of 46.1281, an adjusted R square of 0.4794, and a standard error of 32.4123--virtually the same adjusted R square and standard error, yet a more simple model with a substantially higher F-Ratio and T statistic. CAPTRLQ = 22.9445 + 3.5178 (C%80-90H) (2.805) (6.792) Tables 19 and 20 provide the analysis of variance and residuals for the single independent variable model. Table 21 provides the model's variables for all fifty metro areas. Table A.2 presents select descriptive statistics for each metro area, sorted by CAPTRLQ. I also performed a forward, backward, and stepwise regression on the hypothesized model with the same conclusion that C%80-90H was the only significant variable. 136 Table 18 Proposed Model Regression **** flULTIPLE BEBBESSION *il't't Listeise Deletion of fliuing Data Equation timber 1 Dependent Ua‘idale” CFPTRLO Blockflmber 1. Hethod: Enter 0.80MCEE1181TV8fl.Es.Jfli-I£BILO OBLKLO OHSPLO 001111.13 Wiableu) Entered on Step lumber WIOLLG OBLKLO COENSITV OHSPLO 88LES.T8 l-I-EBILO 0.80.901 «IOU-DON-o mitiple B .73983 8 Sana-e .54734 adjusted R Squrre .47190 Stardwd Error 32.64615 Fhalusis of Ucricnce 0F 31. of Near Squire Elevation 7 54125.77118 7732.25303 Residual 42 44762.38895 10155.77"? F I 7.25508 Simif F I .0000 Uwidalu in the Equation Ucrlable 8 SE 8 Beta T 819 T ”HELLO 1.157855 2.973642 .“995 .389 .69” OBLKLO 1.124010 .856671 .171115 1.312 .1N6 CDENSITV -1.433868 1.326852 -.145613 -1.081 .2860 OHSPLO -3. 160321 4.450977 -.092834 -.710 .4816 SRLESJH -. 129300 .278212 -.060558 -.465 .6445 il-IEBILO 10.297338 55.128212 .031566 .187 .8527 0.80.901 3.539012 .958079 .704257 3.694 .0005 (Constant) 22.273652 43.473242 .512 .6111 End Block Huber 1 811 recreated «rubles entered. 137 Table 19 Single Variable Model Regression eeee HULTIPLE assesssrou est: Listeise Deletion of "lasing Data Equationfltnber1 Depemient Utricble” mo Blookfltaber 1. Hethod: Enter 0.80.“ Uwiwleu) Entered on Step W 1.. W Mitiple B .W B Sqm .4eooa adjusted 8 Square .47943 SW Error 32.41250 Fhalusis of Urriarce [F Sue of Squires Hem Squa-e Evasion 1 48460.78926 48460.78926 Residual 48 50427.37088 1050.57023 F II 46.12808 Signif F I .0000 Ueriwles in the Equatlon Ua-ldrle 8 SE 8 Beta T 819 T 0.80.“ 3.517822 .517954 3111340 6.792 .0000 (Constart) 22.945084 8.179844 2.805 .1372 End Block Huber 1 811 requested wide!“ entered. Table 20 Single Variable Model Residuals, Actual Y, Predicted Y Los thele Nee York Chicago Philodelph Detroit Hashington Boston Houston Rtlanta Riverside! Dallas San Diego I'linnecpoll Niels/8a St. Louis Baltleore Phoenix Oakland Tape/SLP Pittsburgh Seattle Mimi Cleveland Heecrk Derwer Sanrancis Kansas 01t Sacra-ento Sm Jose Cincinnati Norfolk fllleaukee Columbus Fort North San Rntoni Ft. Lauder Portland 0 Bergen/Pas lndlanapol Nee Orlem Charlotte Orlando Salt Lake Rochester Haslwille fleephis Okld'roea 0 Buffalo Greens/Din Dauton E 83828 3s:§::a§ss.s$ssss earaasssraaasssesaasagesssssssassxgarssas333:32233 ‘ 333.3 142. as 114: 3aafiéssssséssssssssssfi:saaééfiaaasssata 138 FEE—2 .47133 .75558 .18241 .0163? .09142 .99156 .54931 :aouz .01331 .23714 .61416 .51791 - .93743 .21238 .o7eaa . 12571 . 59538 . 59622 . 47895 .61031 . 43345 .7699? .45718 .33499 .55441 .50819 .25258 . 19773 .18085 .5281? .72745 .31018 .m910 .9718? .02150 . 11080 139 140 Studen 'zed R iduals--Sin e Ind ent Variable ode Studentized residuals for this single independent variable model, based on the amount of housing structures built in the last deeade (C%80-90H), were then mapped and evaluated for spatial patterns. Only one city had a greater absolute studentized residual than 2.0 (-2.0133), Los Angeles. And only eighteen cities had positive or negative studentized residuals between 1.0 and 1.9, with seven cities being positive. The studentized residuals from the proposed model with the seven independent variables were the same in fifteen of the nineteen above. But the single independent variable model excludes New York and includes Portland, Indianapolis, Greensboro-Winston Salem, and Paterson, NJ. Only six were greater than +1.0 yet less than 1.5 including Sacramento, Kansas City, Memphis, Baltimore, Greensboro and Winston-Salem, and Indianapolis. Only Baltimore is an older, inelastic community with apparently similar success at attracting discounters as found in its downtown redevelopment efforts. The rest of the communities range from medium to high elasticity and have been new growth areas. Some communities, such as Memphis and Sacramento, may also have broad hinterlands given the lack of nearby major centers . Only Portland had a studentized residual between 1.5 and 1.9 at 1.51. Portland has medium elasticity; of interest is Portland's suburbs having the highest suburban square footage per eapita of discount store space among all fifty metro areas and the city having fourth highest. This is especially interesting given the urban growth boundary in Portland attempting to lirrrit suburbanization. However, Portland may have more large-seale supercenters incorporating substantial food sales tint might skew this statistic. Only five were between -1.0 and -1.5 including San Francisco, Oakland, Dallas, Fort Worth, and Paterson , N.J. Possibly the S&L crisis coupled with the oil recession hurt Dallas and Fort Worth although Houston and San Antonio did not deviate. However, both Dallas and Fort Worth have substantial general merchandise sales suggesting that the area has a major competitive supply of department stores as well as discounters. These two 141 citieshadgreaterthan20%housing structure growth in thelastacadeandwereratedas highly or hyper elastic. San Francisco is probably tied to density, assembly costs, architectural preservation, neighborhood business competition, and lack of neighborhood support. Oakland is probably due to its negative reputation and perception as an older, low income minority community. And Paterson is probably due to the density and difficulty of land assembly. Immigrants have kept cities like Paterson at higher than usual densities for its size (Rusk 1993) Also, Bergen County has existing Blue Laws prohibiting Sunday sales, which may prohibit Paterson's success (SCT January 1994, 17). San Francisco, Oakland, and Paterson have zero elasticity rankings. Only five were greater than -l.5 yet less than -2.0 with San Diego, Newark, Boston, Ft. Lauderdale and Nashville. And Los Angeles with low elasticity has a studentized residual of -2.013. With medium elasticity, Nashville might be negative since it has substantial suburban competition with Nashville's suburbs having the lowest population per store of the fifty suburban areas. Although highly elastic, Ft. Lauderdale might be due to an elderly population and the substantial general merchandise competition that exists in the market Newark is inelastic and incorporates substantial urbanization problems of social deprivation. Also inelastic, Boston has heterogeneous neighborhoods that may not be supportive of discount stores given the historieal preservation factor in concert with the density and cost of assembly issues. And San Diego with hyper elasticity may complement the metro areas of San Francisco and Oakland as being too costly and difficult for land assemblage, although San Diego had greater than 20% growth rate in housing structures during the deeade. In addition to the California cost issues, one-third of Los Angeles County's population in 1990 was foreign-bom—-half having arrived in the past decade (Rusk 1993). Note in Chapter 2 that California cities are recently being tested with multi- level discount store prototypes as in the Northeast thereby suggesting this land assembly difficulty. As a further illustration of the difficulty of land assembly in the Los Angeles region, the first Super Kmart Center in Los Angeles County just opened on the site of a 142 former refinery needing extensive environmental remediation and assistance from the State (SCT January 1996, 3). A mapped analysis of these outliers does show some slight regional pattern with California and the Northeast having negative residuals. However, no obvious discernible pattern exists generally across the nation. Insrgmficant Indegndent Variables Since only C%80-90H is significant, the other independent variables in the model representing land availability and investment opportunity are not necessarily indicators of the ability of cities to attract discount store space. This conclusion may be positive, perhaps indicating that cities lave more control over their own destiny by using certain programmatic approaches, regardless of selected apparent negative traits. l. CDENSITY Density is not significant, but the negative sign of the beta coefficient is as hypothesized. Although insignifieant, a review of the lowest twenty-five communities (in terms of CAPT RI.Q) indicates that only five communities are under 4,000 population per square mile whereas nineteen of the highest twenty-five are less than 4,000. Density is also collinear with the significant variable C%80-90H as new development tends to be substantially less dense in elastic cities. Density is a surrogate for land availability at a reasonable price. 2. SALES%TAX The sales tax as a percentage of local revenue was not a signifieant variable in determining the CAPTRLQ, and the sign of the beta coefficient was negative compared to the hypothesized positive relationship. The negative sign could indieate that older, inelastic communities dependent on the sales tax have not attracted discount stores. With the decline of property tax receipts in many older communities, the shift to the sales tax has been considered by many to be a good source of revenue. Although regressive, the sales tax is considered positive especially for those retail projects encouraging sales from customers 143 not living in the city such as tourists, office workers, and inner-ring suburbanites. Furthermore, if market studies show leakage out of the city to suburban stores, then the city is losing the opportunity for retention of city resident sales and the attraction of non- residents. Although this analysis looked at the sales tax as a local source of revenue generation only, further research should clarify whether any other sales tax sharing method exists such as a county or state rebate. Cities might also take an aggressive defensive strategy as do some retailers. The city would be highly susceptible to new stores in the inner-ring suburbs (and even beyond in many markets). Developers and tenants will sometimes locate at a lower priority site in order to act as a defensive strategy if another store is attempting to penetrate or expand in the market. For the city's self-preservation of its retail market, it should look defensively at nearby proposed competition in inner-ring suburbs. Some cities might not prefer to attract large-scale retailing, but they need to perform their benefit—cost analysis in light of the changing competition. Whereas discounters historically competed primarily for department store type merchandise, the evolution to scrambled merchandising makes city stores more susceptible to competition. And the evolution to food stores incorporated into supercenters provides potentially severe competition to city revenues since food stores create substantial sales, usually over $300 per square foot. Cities should also remember that discounters and food stores might serve as anchors for attracting other retailing. So deterioration of general merchandise sales and/or food sales to outlying competition could also have a negative affect on other loeal retailers dependent on the trade generated by the anchors. Even cities not dependent on sales tax revenue need to consider the importance of retailing on other tax creation. And these trends obviously could affect the job opportunities for city residents. This analysis has also not looked at the tax structure of each individual metropolitan market. The intra-urban discrepancies of ratables between city and suburb may be a deterrent to a city attracting retail development. Higher local sales taxes, earnings taxes, business taxes, and property taxes in the central city compared to its suburbs could deter 144 development-this discrepancy could be a signifieant independent variable. The impact of intra-urban differentials should be evaluated in specific community evaluations during further research. 3. HHEBILQ This variable is positive as hypothesized but is insignifieant and collinear with the signifieant variable, C%80—90H. 4. %BLKLQ and %HSPLQ Neither the Black nor Hispanic differentials between cities and suburbs are signifieant variables. The hypothesized negative beta coefficient sign for both variables was only met with the %HSPLQ whereas Blacks were positively related. However, the insigrrifieance of these variables could be due to the level of aggregation. Case studies should determine spatial loeational patterns of store loeations at the individual market level. Looking at the descriptive statistics of percentage of Black population in the city, the lowest twenty-five cities are all greater than 10% Black whereas eight of the top twenty-five cities are less than 10% Black. Many of the elastic cities performing well have substantial Black populations, but their Southern location in newer growth communities may represent a different demographic nature of Black households compared to less elastic communities. This finding suggests that discounters may serve minority markets where land is available and neighborhood life cycles are adequate. Further research at the individual market level should focus on trade area differential in cities of different elasticity. No apparent pattern exists for Hispanic population. 5. VIOLLQ This variable is insignificant and the positive sign of the beta coefficient is not as hypothesized. Although the city: suburb ratio is insignificant, only seven of the lowest twenty-five cities had less than 1,500 violent crimes per 100,000 population compared to fourteen of the highest twenty-five. Again, this variable is collinear with C%80—90H. 145 D_esgg'ptjve M Caveats As stated previously, the data are from 1992 and the discounters are testing prototypes for multi-level stores in very dense and expensive land markets such as New York, Los Angeles, and San Francisco. Furthermore, with suburban saturation, discounters and developers have been active recently in public-private partnerships that may reflect improvement in low eapture cities. For example, this analysis incorporated approximately 210,000 square feet (Kmart and Sam's) in a new 500,000 foot power center in St. Louis that opened in Fall 1991 and St. Louis has recently announced a 200,000 square foot convenience center with a 100,000 foot regional discount department store, loeated in the minority community. Both of these projects require tax increment financing and extensive land assembly assistance by the city. Therefore, more recent data nright show some improvement in those cities having low CAPTRLQ. It should be emphasized that these findings merely show the existing spatial distribution of space in 1992 and do not necessarily indieate a lack of demand in cities with low CAPTRLQ, which might actually be an indicator of unmet demand. The data also do not indieate whether non-discount store strategies ean be successful in cities. The findings provide a foundation for planners to compare their community with others and to develop their own retail strategies including discount stores. High CAPI'RLQ cities are theoretieally presumed to be importing non-resident sales with low CAPT RLQ cities exporting sales to the suburbs. It is possible that city resident shoppers are not shopping outside the city limits and are shopping at alternative shopping loeations such as department stores or smaller business districts and stores. Furthermore, this analysis only looks at square footage relationships between cities and suburbs; it is possible that sales per square foot are substantially different in central city stores. Descriptive data indieate that cities have a median of 1.2 square feet of discount department store space per eapita compared to 2.6 in the suburbs. Minimum space per eapita is zero in six cities with a maximum of 5.0 feet per eapita in the city of Orlando. 146 Minimum suburban space is 0.9 feet per eapita in Newark and a maximrun of 4.4 feet in suburban Portland Population per store varies widely in the central cities from a minimum of 18,300 in Orlando to 871,350 in Los Angeles with a median of 74,561. Suburban population per store ranges from 18,393 in Nashville to 87,970 in San Francisco with a median of 31,906. CAPTRLQ ranges from 0.0 in six cities to 189.94 in Orlando with a median of 63.85 and an average of 68.96 and a standard deviation of 45. Only 13 of the 50 cities have a CAPTRLQ greater than 100. Not surprisingly, regional patterns exist for CAPTRLQ with the Midwest and Northeast having only three communities higher than 100.0-Kansas City, Indianapolis, and Columbus-all communities that expanded their boundaries in recent deeades and have suburban-style growth within their boundaries. The signifieant independent variable, C%80-90H, coincides with this pattern. The only cities with less than 5% housing growth are all loeated in the Midwest and Northeast-11 communities. Six communities have no discount department store space including San Francisco, Oakland, Ft. Lauderdale, Newark, Paterson, and Boston. At the other end of the spectrum, only Sacramento and Orlando have greater than 150 CAPTRLQ, but under 200. General Memhflrdise Sales versus QAEIIM I have prepared a GMCAPTRLQ variable similar to the dependent variable in Table A.2 representing the general merchandise sales capture quotient in the community. These 1991 year-end sales are for SIC Major Group 53, which includes retail stores that sell a number of lines of merchandise, such as dry goods, apparel and accessories, furniture and home furnishings, housewares, hardware, and food. This variable includes conventional department stores, discount department stores, limited-price variety stores, and miscellaneous general merchandise stores and catalog showrooms, although catalog and mail-order operations are not included. Data from the 1992 ”Current Business Reports for Retail Trade' show that approximately 45% of the national general merchandise sales are in 147 conventional or national chain department stores, 39% in discount stores, 4% in variety stores, and 21% in miscellaneous general merchandise stores. These should not be confused with individual stores for apparel, furnishings, household appliances, and other shoppers goods. I have not evaluated the level of competition within the city or the metro market to determine the degree to which a low CAPTRLQ for discount department stores is negative for the city's retail equilibrium. For example, San Francisco appears to have a vibrant overall retail commrrrrity compared to Oakland, yet both have zero CAPTRLQ. The same distinction could conceivably be made for Boston compared to Newark. Even Minneapolis-St. Paul, the home of Target Stores, has a low CAPTRLQ, yet they have historieally had aggressive downtown retailing programs. Note that many of the cities with low CAPTRLQ have fairly high GMCAPTRLQ, suggesting that either the discount stores have very high sales volumes per square foot, or most likely, that these cities have substantial department store space in their communities attracting suburbanites and visitors. For example, Minneapolis-St. Paul, Cincinnati, Ft. Lauderdale, Boston, San Francisco and others all have much higher GMCAPTRLQ than their discount store space eapture. Multi-collinearity between these two variables exists but analysis of the data suggests that some cities not performing well in discount store space are attracting general merchandise sales. Orlando is the highest in both categories. Each city should determine its own level of retail saturation and the necessity to attract discount stores as part of its overall retail supply. The degree of leakage from each city ean be easily determined through market research. Degree of Elasticig Table A.2 also includes Rusk's relative elasticity eategories representing the combined effect of a city's density in 1950 and the degree of city limit expansion between 1950 and 1990. ”Each city's initial density and degree of boundary expansion (by 148 percentage) are ranked against those of all other cities in its group. A city's relative rankings (organized by decile) for the two key characteristics (initial density and boundary expansion) are multiplied together to produce a composite elasticity score." It should be noted that the degree of elasticity is inversely proportional to metro area size. The degree of elasticity does have a correlation with CAPI‘RLQ. I included elasticity as an independent variable in the hypothesized model and the single variable model, however, and elasticity was neither significant nor an enhancer of the adjusted R square. Elasticity ranges from 0-4 with 4 being hyper-elastic. Of the lowest twenty-five communities (with CAPI'RLQs of 0.0 to 63.0), eighteen are ranked as Zero Elasticity with four having low elasticity, two medium, and one at high elasticity. Of the highest twenty- five, only two have zero elasticity, three are low, seven are medium, seven are high, and six are deemed hyper-elastic. Tire degree of elasticity is also collinear with the signifieant independent variable, C%80-90H. However, it should not be assumed that cities with high CAPTRLQ have discount facilities serving their older neighborhoods, since all the stores could be at the new growth edge of the city. These cities may incorporate the same limited supply of space in the old, "inelastic” portion of their cities as do the inelastic slow growth communities. Rusk (1993) suggests that annexation or consolidation with urban counties is the primary solution to the disparity existing in the inelastic cities. While that would undoubtedly be positive from a fiseal standpoint, older areas may still be deficient from the perspective of having convenient discount store shopping, which could be a deterrent to neighborhood preservation efforts. Case studies of elastic cities should evaluate whether the older portions of the community are attracting development as well as the new growth portions of the city. Men: Only one variable, percentage of new housing built during the last deeade, was deemed significant with an adjusted R square of 0.47 but with a high standard error. Thus, 149 new development apparently represents both life cycle and land availability considerations. Only one city had a studentized residual just over 2.0 with eighteen of the fifty cities between 1.0 and 1.9. Further analysis of the model incorporating measures of elasticity and other general merchandise sales (as an indicator of competition) do not provide model enhancement. The next chapter provides conclusions and recommendations for future research. CHAPTER7 SUMMARY AND CONCLUSIONS Perhaps the failure of the model should be viewed positively for cities since the model's deterministic framework suggested that certain variables (maybe beyond a city's control) would dictate the availability of discount store space. Although revisions to the model did not consider elasticity as good an explanatory variable as the amount of housing built during the last decade, a review of cities performing better tend to show elasticity being positively correlated to capture rates. Yet, it is apparent that the rate of capture even varies greatly among the twenty inelastic cities, ranging from zero to 82.5 CAPTRLQ. Retail Difficulties in Hetemgeneous Cig Markets Retailers and developers tend to develop prototype formats that meet the mass market and the typical homogeneous suburban site location. Deviance from the formula prototype causes less efficiency and must be weighed against market potential and return on investment. From a discount store standpoint, only recently have discounters started to micro-market their merchandise and marketing techniques to meet the local market, even at the store unit level. Economies of scale historically have been formulated on standard prototype store sizes, interior layouts, merchandising, and pricing. Suburban trade areas for individual stores have been generally homogeneous because of the nature of suburban zoning and development patterns. Usually, residential life cycles are homogeneous and the demographics of trade areas are somewhat homogeneous from the perspectives of class, occupations, age, race, income, and other psychographics. In the central city, however, neighborhoods within potential discount store trade areas are 150 151 extremely heterogeneous from the above perspectives. Thus, merchandising becomes more complex. Class and racial bias starts to have a potential negative effect on shopping patterns as certain higher economic social groups will shun stores that have either a perceived or real clientele of lower class. Studies have shown that shoppers prefer to have salespeople who are of their same background, a potential deterrent with certain municipal job program efforts. "Studies on retail sales personnel show that customers react more favorably if salespeople are perceived as being like themselves. This is particularly true of the racial similarity of the customer and the salesperson. As product complexity increases, consumers place greater emphasis on the salesperson's product expertise, empathy, trustworthiness, personal appearance, credibility, and professionalism (Berman and Evans 1989).“ City neighborhoods are very diverse although segregated economically and racially in either pockets or large portions of the community, usually in a concentric or sectoral zonal concentration. Yet, discount store trade areas may extend 3-5 miles incorporating a truly heterogeneous population. Recognizing that racial and economic bias exists, it is possible thatastorecould becornetlre shopping locationforonlyonesegrnentofalocal tradearea thereby diminishing the market potential and economic success of the site. For example, a store in a mixed income minority trade area might become a shopping location for only lower income minorities whereas upper income minorities might perceive those shoppers negatively (class bias) and shop at alternative suburban units having a more homogeneous shopping population. Yet, the store located at the site in anticipation of the higher income shopper being a part of the trade area. Add to this example racial bias and discrimination in mixed racial communities. The same concepts occur in white communities in city locations as yuppies and/or higher income whites may not prefer not to shop with lower income whites, at a certain percentage of the shopping trade. The difficulty for the merchant in these heterogeneous trade areas is to provide the merchandise for individual niches, a very complex ordeal in such diverse markets. 152 A mixed income and racial consumer profile can be a psychological deterrent to store location decision-makers, who do not have experience in dealing with diverse markets or may have had negative experience with stores in such markets. Where shopping patterns of city residents to suburban locations have been established due to an historical lack of facilities in the city, it is difficult to alter shopping behavior. And if those alternative suburban shopping areas were perceived as meeting shopper needs, even a more convenient new center in the city may not be considered a legitimate shopping alternative, especially if the shoppers are more heterogeneous, the neighborhood is not perceived as similar to their own, and if the perception/reality of crime is more prevalent at the city site. Given the difficulties of creating a large-scale retail agglomeration at the city site with possibly only the ability to have a small center, the suburban location may still be preferred because of multi-purpose shopping opportunities in nearby regional malls or other retail shopping centers in the vicinity (Ghosh 1986). Another major deterrence to the discount store decision-maker may be the lifetime return on investment for both the property (under either a lease or ownership format) and the commitment to operations for the unit compared to suburban locations where growth is stable or growing. Central city population and income declines vary by city and neighborhood, but certainly the reality of central city decline over the last forty years is negative in reality. Decisions are preferred for at least a twenty-year life cycle and many investnrents are made on the basis that the underlying value of the land in the future will have value, even if the store is closed or relocated. The life cycle concept pertains to all real estate and certainly the city has shown significant decline processes with debatable likelihood of stability or rejuvenation. Most central cities have a higher percentage of multi- family, rental housing than their suburbs, which presents a likelihood that cities will continue to decline as ownership is not a reasonable possibility. Furthermore, single- family and homeownership areas tend to lend stability to the environs for commercial use 153 as properties tend to be better maintained and owners tend to be older, higher income earners with children. Although constraints exist, cities tend to have a severe shortage of modern retail space and merchandise offerings; furthermore, the population density is extensive creating opportrrnities for convenience shopping. Many local governments are willing to provide incentives and eliminate the negative barriers to redevelopment. Local governments are very active with neighborhood organimtions in preservation and revitalization of their neighborhoods, creating more stability for the area. Many city neighborhoods have attracted young households with growing incomes and have attracted new immigrants to their neighborhoods. Less suburban opportunities exist due to overbuilding or slowdown in growth, yet many underserved city neighborhoods exist that can be legitimate investment options. Also, many discounters previously closed city locations from the 1960s or have obsolete facilities immediately adjacent to city locations in inner-ring suburbs; city locations may be a legitimate alternative location. Mgket Research Difficulties Perforrrring market analysis for proposed developments in central cities is complex because of the inability to find cornparables and analogs. Shoppers surveys for stores or any kind of demand survey analysis may be biased based on the nature of the existing supply. Given the fact that a substantial majority of the shopping centers and discount stores are located in the suburbs, shoppers surveys provide statistics potentially biased to the suburbs. Thus, traditional market analysts assuming that certain types of shoppers will only shop at proposed central city markets may underestimate the market potential if the demographic data for the city is different from the suburbs, which they are. While certainly income is a key consideration, other factors of age, occupation, household size, and related socioeconomic factors may create new opportunities for discount operations. At one point, many food store chains were hesitant to upgrade their facilities in central cities (e.g. new, remodel, expansion, relocation) because their existing relatively poor facilities were 154 not performing adequately. However, many food stores did improve their facilities and have significantly positive returns on investment. Similarly, new residential development or gut rehabilitation started in the central cities in the 19703 without many successful prototypes nationally and quite a few developers have penetrated this niche market. Yet, standard market analysis techniques were not really useful because proposed central city development frequently goes against the existing market trends or comparables. This market problem also becomes an appraisal problem for the shopping center as an investment due to a lack of comparables either for finding properties that have been developed/sold in the city or for finding comparable rents. The new shopping center will have rents that are typically substantially higher than the unplanned strip ribbon commercial spaces in haphazard available space along major corridors, especially in an older city. Proposed facilities may even be in the midst of vacant and abandoned commercial ribbon space making it difficult to estimate the market/appraised value given the seemingly saturated commercial climate of the vicinity (albeit functionally and econonrically obsolete). Obviously, the best way to overcome this problem is to have signed leases from AAA tenants, but remaining speculative space in a center becomes difficult to analyze. In an average midwestem community, rents for new strip center speculative small tenant space may be $12-15 per square foot whereas adjacent obsolete space (without parking as well) may rent for $25. This discrepancy may also be a problem to attract national tenants-not only are they concerned about the market but they are concerned about the impact of the obsolete space. And if the market analyst/appraiser is having a problem with the project, the same real and perceived concerns exist for the permanent lender who needs to be concerned about the project's ability to finance any debt and the project's collateral position as an investment if any problems arise in management/ownership. Finally, all of these decisions are being evaluated in light of other opportunities that may be possible for investment in other projects, financial instruments, or leasing opportrrnities. 155 What might a local city planner do to consider how to attract retail developers, assuming such development meets the city's goals and comprehensive plan? trim—SW Unlike suburban-type areas (including in some elastic central cities), large sites are frequently not existing or apparent in the central city. It is very difficult for city planners to take a developer/tenant to an area that needs land assembly (with existing residential or commercial/industrial uses) and identify the area as a potential site for redevelopment in less than two years. Developers and tenants are looking for sites that are relatively easy to develop in a timely manner to meet their capital needs for growth. Developers/tenants are acting in good faith that the city can meet any partnership obligations and commitments. Although legitimate bureaucracy and public processes exist in suburban settings (primarily regarding rezoning), they are rarely as complex as a public-private partnership and the related political processes, especially when public financing is involved. Thus, transaction costs and risks are usually higher in inelastic central city developments. Although many have critiqued the urban renewal program that existed from 1949- 1974, the ability to have sites available for development according to a comprehensive project plan with funds set aside for the project was a positive attempt to reduce transaction costs and eliminate the many previously cited barriers to redevelopment. Many cities are unwilling to assemble sites without a developer in hand, thereby potentially limiting the number of developers or tenants adequately interested to pursue higher risk. Thus, it is important for the city to identify sites and to show the retail community that the city government has the ability to implement the public-private partnership expeditiously. Furthermore, it is important for the city to identify realistic sites for development and encourage sites that are in relatively positive trade areas versus showing sites in severely abandoned areas, which may have inadequate funding for total revitalization in order to eliminate negative exterrralities. 156 Prior to a detailed market study, the redevelopment city planner could implement these steps within two months with a couple of staff in markets under several million population. 1. Inventory the existing and proposed supply of major retail anchors in the metro area or at least in the nearby competitive suburban area as well as the central city. The inventory should focus on discounters, category killers, food stores, hardware, drug, and department stores. Specific data should include the competitive characteristics of each store such as square footage, age of facility, strength of the center and the adjacent environs, and ability to expand the center. Evaluating a map of these stores will provide an excellent visual perspective of potential spatial trade area opportunities for city development. Preliminary trade area concentric circles at l, 3, and 5 miles can assist in evaluation. 2. Identify potential sites through a combination of field research (i.e. drive along major corridors) and base maps with building outlines by parcel (in order to see large land areas potentially capable of assemblage). Aerials can assist also. a. Look initially at major intersections of highways and major arterial streets, then along those corridors close to the intersection. b. Identify abandoned or underutilized major industrial/commercial facilities that could be demolished and reused. Many central cities have single-unit vacated or obsolete mass merchandisers such as Sears or department stores built from the 19308 through the 1950s that provide an opportunity for discount store locations. Major free-standing department stores built in the early fifties are well located on major arteries, but they are frequently multi-story buildings in the 300,000 square foot range, too big and functionally obsolete typically for retrofit to a discount store. As described herein, building reuse may be feasible in very dense markets where no land opportunity exists. Perhaps these buildings could be demolished; they frequently have substantial land available for parking, although the site may need to be 157 expanded for modern standards and to allow for larger developments such as power centers. c. It is important that cities consider sites that might require either commercial or residential relocation. Good commercial sites that could be physically assembled at some reasonable cost should be identified. The politics of land assembly needs to be considered, but early stages should look at all options. d. If pertinent to the nature of the historical or existing land use, prepare a Phase 1 Environmental Review for potentially good development sites. 3. Prepare a preliminary land assembly cost estimate including costs of acquisition, relocation, demolition including environmental clean-up, and site preparation. 4. Estimate the development capacity of the site and an estimated sales price for land disposition. Estimate any land writedown costs (difference between land assembly and land disposition cost showing the necessity for subsidy). 5. Identify the financing tools available at the local level such as CDBG, tax increment, tax abatement, enterprise zone, state funds, etc. Determine the appropriate planning and public processes including any aspects of blighting, land assembly, and developer solicitation/bidding. 6. Meet individually with key metro area shopping center developers and major commercial Realtors to discuss the general market potential and what it would take to attract major anchors to the city. Without being too specific at this stage, identify general site locations to deterrrrine their reaction. Determine what kind of market information they would prefer and what kind of marketing package would be necessary to take to the appropriate anchors. 7. Review retail sales trends in the city and adjacent areas to determine the importance of retailing to the community. 8. Prepare preliminary market potential estimates for each site using census data or ordering national market data by trade area. 158 9. Prepare telephone market research of city and suburban residents focusing on trade area shopping patterns to determine the strength of the city retail market and the potential leakage from city residents to suburban stores. 10. Determine the appropriate political course of action regarding the general issue and particular sites. Based on reqrrired public processes for developer solicitation and on the desired public/political relations, consider holding a developers' workshop to solicit proposals for selected sites. Or work with a limited number of developers who have expressed interest based on prior contacts. These steps would provide a relatively cost-effective means to quickly assess the potential for such development and would lay the foundation for further evaluation, especially pertaining to benefit-cost analysis for retail proposals. Future Research These research findings certainly suggest that land availability is a key variable. The ability for a city to attract discount store space may also be a function of the city's direct efforts at assisting in the land assembly process, especially in older, inelastic built-up communities. Substantial research opportunities exist. 1. Case studies on a longitudinal basis should be prepared of both successful and unsuccessful discount department store developments in central cities to serve as analogs. These analogs will be vital for cities to attract developers/tenants and to assist in public deliberations at the neighborhood and local legislative level. Perceptions must be complemented with facts. Many developers and retailers have negative perceptions of central city retailing, frequently based on the reality of selected failures in the past. And many neighborhoods are rightly concerned about both the impact of new retail development on existing merchants and the impact of large-scale stores on their overall environs. Cities must be vigilant in providing detailed information regarding successful and unsuccessful cases. Studies should incorporate details about the financing and political process, market 159 penetration, impact on small stores and adjacent neighborhoods, job creation and related city programs, fiscal impacts, and any effects as a catalyst for additional development. These analogs should be prepared by surveying community development/planning agencies, shopping center developers, retail market analysts, and discount department store chains to determine why some cities have either attracted or not attracted discounters. Questions for the community development/planning staff should include: * Number and name of discount stores in city and perceived need * Age, condition, and location of stores * Why city is doing well or not in discounting * History of discount store projects and city assistance * Efforts to encourage developers/anchors to city * Eminent domain powers and willingness to use * Frnancial tools available to assist development * Availability of sites " Interest in encouraging such development Questions for the private sector should include the following aspects: * Experience with central city developments " Locations of central city developments * Demographic and market issues * Operating issues such as return on investnrerrt, merchandising, employees, operating costs, pilferage/security, and credit. * Site location issues * Rents, taxes, and costs compared to suburban locations * Political environment "‘ Required financial incentives for existing locations * Requirements to locate more in central cities, including minority areas 160 Researchers should consider results from an historical perspective. Some early discounters did have stores in cities initially as their business evolved but many were unsuccessful. In evaluating any negative responses of the private sector to central city investment, researchers should consider the historical realities and perceptions of central city discount store decision-makers including: * Was being in the central city the problem or were discount operations not adequately developed yet as a store type? * Were the locations of stores adequate and were the adjacent commercial/residential areas adequately upgraded? * Was the operation too homogenized catering to its evolving suburban mass merchandise experience, without considering the heterogeneous nature of city neighborhoods? * Were initial operations developed in central cities during the major period of city population decline thereby creating a volatile market to penetrate? Did the riots in the late sixties have real or perceived negative irrrpact on investment decisions? * Are central cities now less volatile demographically and have they stabilized into a lower income and racial/ethnic niche opportunity in selected city locations? * Does a central city limit bias occur regardless of the demographics and even the growth opportunities in the community? * Have the local government's typically required tools of redevelopment, especially eminent domain, and neighborhood preservation been institutionalized at the local level today corrrpared to the past in order to give confidence for private sector investment? 2. The significance of C%80-90H, or the amount of housing built over the last decade, has been deemed a surrogate for land availability at a reasonable price. Case studies of inelastic communities performing better than others are hypothesized to show aggressive partnership endeavors by the city to make land available. Through the case studies, research should focus on the following questions: 161 * Is land availability at a good location for a reasonable price the primary variable, or is the C%80-90H variable primarily an indicator of the life cycle process, which may preclude older areas from attracting retail development ? * How aggressive have cities been in attracting developers/discounters to older neighborhoods; has land been either made available or identified including appropriate public incentives ? *Do cities with high CAPT RLQ have similar trade areas as those not performing as well including degree of racial, ethnic, economic, transportation, and land use characteristics ? 3. Another research topic is to look at the degree of competition in each market by store size and by company to see any trends and potentials for future development. Regional discounter trends versus national store spatial strategies should be evaluated. This research should emphasize the potential vulnerability of city stores to restructuring trends in the industry, especially towards supercenters or larger-sized discount stores. 4. Given the nature of central city demographics, research on personal consumption expenditures by lower income, older age groups, renters, and minorities should be pursued. In the last decade in many city markets, food stores finally decided to expand their supply and meet the demands in heterogeneous city neighborhoods. They should serve as effective analogs to developers and tenants of the retail opportunities in the central city, which appear to be understored in this analysis. Also, further research into the evolving micro-marketing trends of large-scale retailers should be pursued to determine their effectiveness at meeting the needs of heterogeneous, non-standardized trade areas. 5. Research should evaluate any additional costs that may be required in central city locations, especially in older portions of those cities. Both capital and operating cost differentials should be evaluated. Cities would then have a better understanding of the possible competitive disadvantage of a city location from the developer and tenant perspective. Furthermore, individual metro market evaluation of the intra-urban 162 competitive environment should include tax differentials and impacts of required city requirements for any assistance such as extensive architecture or landscaping, irnprovements/exactions, minority/female employment goals, red tape, etc. compared to alternative suburban sites. Mam Every city needs to determine the desirability of encouraging certain land uses including discount store retailing as part of its retail offering. Discount stores have become the mainstay for the American consumer and are an integral part of the national and local economy. Benefit-cost analysis and development impact analysis are effective tools to deliberate the role of retailing in a community generally and for site specific locations. Retailing is definitely a vital component of neighborhood preservation efforts with cities needing to continue the modernization process for extended neighborhood life cycles. Furthermore, retailing provides entry level job opportunities and substantial tax creation. Retail demand and opportunities vary by community and by sub-areas within each city. This research suggests that discount stores are an important part of retailing, but not every central city needs discount stores—especially if department stores and other smaller establishments can meet the retail needs of the community. But discount stores are such an integral part of total retailing that it is very possible that many city residents are shopping in the suburbs to the detriment of the city's economy. It is imperative that city planners perform to their utmost ability to use the local financial, legal, and administrative tools available to encourage positive development for the community. For many communities, discount store development is positive and can be an economic anchor for many land use and development opportunities. GLOSSARY OF TECHNICAL TERMS The following terms are used by the retail industry with several being synonymous as used loosely in the trade. big box retailer. A nickname, referring to the physical characteristics of the tenant building, for large scale value oriented retailers including both category killers and discount stores. category killers. Stores that specialize in one retail segment, offering a wide breadth of inventory only in that segment, with high-volume sales generated through extensive advertising and marketing. Such stores include The Sports Authority, Office Depot, Home Depot, Supermarket of Shoes, Circuit City, and Toys R Us. These stores are either freestanding or frequently part of power centers-typically in the 20,000 to 60,000 square foot range. Also synonymous with promotional anchors and deep discounters. convenience goods. Those goods needed immediately and often and which are purchased where it is most convenient for the shopper. Also known as lower order goods, these goods include food stores, liquor, drugs, sundries. conventional department store. Establishments meeting the criteria of a dcpartrrrent store and usually provide check-out service and customer assistance within each department, such as Hudson's, Rich's, Lord & Taylor, and Nordstroms, typically over 100,000 square feet. department store. Generally, a large retail business that handles an extensive assortment (width and depth) of goods and services and is organized into separate departments for purposes of buying, promotion, customer service, and control. A department store is a departmentalized retail establishment that sells many lines of merchandise including men's and women's clothing and accessories, piece goods, small wares and home furnishings. Department stores are a sub—category of the general merchandise classification including conventional, national chain, and discount department stores having more than 50 employees, and sales of apparel and sot goods combined amounting to 20% or more of total sales. deep discounter. See category killer. discount or off price store. A limited service retail establishment utilizing many self- service techniques to sell hardgoods, health & beauty aids, apparel and other soft goods, and other general merchandise with centralized check-out service. It operates at uniquely low margins, has a minimum annual volume of $1 million, and has at least 10,000 square feet of total space. Such stores include Marshall's, Burlington Coat, and Ross. factory outlets. Manufactured-owned store that sell manufacturer's closeouts, canceled orders, discontinued merchandise and irregular goods. They are now located in outlet mall shopping centers and compete with other value oriented merchants. 163 full 164 full line discount store. A departrnentalized retail establishment utilizing many self - service techniques to sell hardgoods, health & beauty aids, apparel and other soft goods, and other general merchandise with centralized check-out service resembling a department store. It operates at uniquely low margins, has a minimum annual volume of $1 million, and has at least 10, 000 square feet of total space. Conveys the rmage of a high- -,volume fast turnover outlet selling a variety of merchandise for less than conventional prices. Such stores include Wal-Mart, Kmart, Target, and Venture. Typically sized over 40,000 square feet, modern units are over 70,000 feet with many now over 100,000 square feet. general merchandise sales. SIC Major Group 53 including sales by department stores, variety stores, and miscellaneous shoppers goods stores. gross leasable area (GLA). Total floor area for the tenant's occupancy and exclusive use in a shopping center--including basements, mezzanines, and upper floors-- expressed in square feet and measured from the centerline of joint partitions and from outside wall fences. It is all that area on which tenants pay rent, including sales areas and integral stock areas. higher order goods. See shoppers goods. hypermarket. Similar to a supercenter, only larger. A hyperrnarket ranges from 200,000 to 300,000 square feet, whereas a supercenter is generally 100,000 to 200,000 square feet. Hypermarkets include a complete supermarket as well as specialty departments such as a bakery, deli, and other services. Most have an extensive general merchandise offering, but a narrower presentation than discount stores, focusing on quick-moving items. See supercenter. lower order goods. See convenience goods. national mass merchandiscr. Establishments meeting the criteria of a department store, and individualized departmental check-out service and customer assistance, but frequently have catalog order service and corporate branded goods such as Sears and J. C. Penney. power shopping center. A power center is defined by the lntemational Council of Shopping Centers ICSC) as a shopping complex of 200,000 to 400,000 square feet or more, at least 75% of which is leased to leading deep discounters, or category killers that are regionally dominant like Circuit City or Kids 'R' Us. Typical centers have around 35% anchor tenants. These centers usually have three to five promotioml anchors selling both hard and soft goods. promotional anchors. See category killer. shoppers goods. Those goods that shoppers will take more care and spend greater effort to purchase. Also known as higher order goods or comparison goods. Shoppers goods (GAFO Merchandise) are defined as general merchandise (i.e.department/discount/variety stores), apparel, furniture, and other miscellaneous shoppers goods (i.e. books, stationery, sporting goods, etc.). shopping center. A group of architecturally unified commercial establishments built on a site which is planned, developed, owned, and managed as an operating unit related in its location, size, and type of shops to the trade area that the unit serves. The unit provides on-site parking in definite relationship to the types and total size of the stores. 165 supercenter. A supercenter is a combination of a complete, f ull-line discount store with a supermarket. The units generally range from 100,000 to 200,000 square feet. hyperrnarkets ranging from 200,000 to 300,000 square feet. Hypermarkets include a complete supermarket as well as specialty departments such as a bakery, deli, and other services. Most have an extensive general merchandise offering, but a narrower presentation than discount stores, focusing on quick-moving iterrrs. traditional department store. See conventional department store. warehouse or wholesale club. Also known as warehouse store or warehouse membership club. A discount retailer that offers a moderate number of food items in a no-frills setting. Straddles the line between wholesale and retail by appealing to price conscious customers who are required to be members in order to shop there. Examples includes Sam's and Price Costco. ' APPENDICES APPENDIX TABLE A] TOP 54 DISCOUNT STORE CHAINS IN SALES, 1992 Table A.1 T 54DiscountStoreChailsInSelcs. 1992 ($50 Million In Salas) Name&Locetioe 166 SeieelStore {1 HT . S) TotalSaies (MillionsS) 1. Wei-Mart Beanstalk AR 2.1{rlsrt Troym 3. Target WW 4. Vehe Merchants Milwaukee WI 5. M Thrm Acres Grand M G.Aue RockyHiaCT 7. More“ Wakefield MA 8.Celrlor NonvakC'I' 9. Bredbes BMW 10. Green WI 11. Venture O'Faaon MO raun- CantonMA 13. Bears Henderson NC 14. Couolrletad Store Coho-aha 0H 15. W (Vets Ch) Columbus OH 16. J Seem NJ 17. Lech-ere Woburu MA 18. Sam's Phoenix AZ 19. M's 3. Wheel New Casio PA 20.Fedco Santa Fe Springs CA 21.?” Orr-alasNE ”PERM DominguarCA 23. 8‘3 Cinch-rad OH 24. Clever Philadelphia PA 1.8) 2.282 517 532 74 419 136 127 111 154 217 541 75 rm 21 18.19 11.10 20.10 0.82 39.730 7.703 6188 15.662 14.409 15.649 1 8.473 11.039 6.470 1.717 12.000 7.926 33.478 36.667 7.778 75.000 3.611 2.750 7 1.429 16.200 34.21!) 10.390 4.363 2.940 2.650 2.551 2.130 LmO 1 .737 1.718 1,700 Tabb A1 (cont'd.) 25. Fort-not! Watery NY “Imam-ace) annPA 27. BM Eugene OR 2am ElinNC 21m MIN 30. Shh Mart Jacksonville FL 31. th's Satan MA 32. A- 8 Hope Cranberiand R! 33. Qeelty Shre- Muategon MI 34. Girl Akron OH 35.A1eo(Deckwel8bres) AbileneKS 36. Hart Stores Columbus OH 37. Jack's Dbceefi Stores Quincy IL 38. Grandpa's Bridget» MO gels: Bay WI 40. Rhine GerDMen Ornalra NE 41.” Dodge CityKS 42. SM'I Cincinnati OH “Anderson's Mar-nee OH 44. 0.1. Joe's Wflronvifle OR 48. Stuart's Franklin MA “machined Honolulu!!! 4‘7. ModelSho Work! East Meadow NF 48 V. Lee-en's Chair-rad OH 49. Ge- “Havel Honohrlu HI 50.13am WorcesterMA 167 41 1% 51 17 10 21 31 14 21 K1000 14.769 8.537 7.88 1 .474 5.451 10.385 3.600 23.000 2.813 7. 143 11.647 17.000 6.875 7.810 4.89 18.250 24.333 9.857 6.429 43.333 3.108 13.286 18.000 12.143 s w 8 E§§§§§§§ 164 11) 146 146 138 135 130 115 168 Tabb A.l (cont‘d.) 51. BergehTm USA. 51 Birmingham AL 52. Apu 3 Warwid RI 53. our We 10 Mt. ”cam: 54. Strekh Group 10 Wichita Falls TX Total 9.015 19.000 5.1!!) 5.11!) 11.778 55 50 115.18) Source: DM, June 1993 Note: Someerrorincalculationby thesourceappenrs tohavebeenmade. Total sales is the figure used previously for all 9,562 stores. The analyst apparently attempted to estimate sales of the top 54 firms to include all sales. Although incorrect, the information is a close approximation since companies with less than $50 million in sales have a minute portion ofthetotal. 'I'hedataaremlybeingusedhereinfordesaiptivepurposes. APPENDD( TABLE A.2 SELECT DATA SORTED BY DEPENDENT VARIABLE 169 170 171 REFERENCES CITED REFERENCES CITED Achabal, Dale D. ,Wilpen L. Gorr, and Vijay Mahajan. " MULTILOC: A Multiple Store Location Decision Model. " Wag, 58, 2 (1982): 5- 25. Adams, John S. "Presidential Address: The Meaning of Housing rn America." Annalsgf theAsscciaunncLAmericanflecstanhetsMM) g(1984) 515- 526 Anderson, Fred B. and Dale R. Funderburk. "Real Estate Valuation: A Pedagogical Approach" IhcAunraisallcumal (April 1989): 221-232 Anderson, Patricia M. "Association of Shopping Center Anchors with Performance of a Nonanchor Specialty Chain's Stores. " lQnmainLRetn'ijng, 61, 2 (1985): 61-74. Anglin, Kenneth A., James W. Gentry, and Jeffrey J. Stoltman. "Shopping Choices: The Case of Mall Choice." Advances in Consumer Research, 18 (1991): 434-440. Applebaum, William. W. Sponsored by Super Market Institute, Inc. Reading, MA.: Addison-Wesley Publishing, 1968. Baerwald, Thomas J. "Changing Sales Patterns in Major American Metropolises, 1963- 1982." Why , 10,4 (1989): 355-374. Bates, Dr. Albert D. "The Extended Specialty Store: A Strategic Opportunity for the 1990s. " W, 65, 3 (1989): 379-387. Bellenger, Danny N, Barnett A. Greenberg, and Dan H. Robertson. "Shopping Center Patronage Motives." W. 53, 2, (1977): 29- 38. Berman, Barry and Joel R. Evans. W. New York: Macmillan Publishing Company, 1989. Berry, Brian J. and John B. Parr. ' ' -- Appiignjgns. Englewood Cliffs: Prentice Hall, 1988. Bier, Thomas. "The Prospect of Accelerated Housing Filtering 1n the 19903: The Cleveland Case." In Bu.“- 7 i a .. , . ._ Black, William C. "Choice-set Definition in Patronage Modeling." Wing, 60, 2 (1984): 63-85. Bloch, Peter H., Nancy M. Ridgway, and Jarrres E. Nelson. "Leisure and the Shopping Mall " W 180990445452 Bluestone, Barry, Patricia Hanna, Sarah Kuhn, and Laura Moore. W Boston: Auburn House Publishing, 1981. 172 173 Boume, L. S. "The Roepke Lecture in Economic Geography--Recycling Urban Systems and Metropolitan Areas: A Geographical Agenda for the 19905 and Beyond. " Eggngnic W 67 (3) (July 1991). 185 209. Bradbury, Katharine L., Anthony Downs, and Kenneth A. Small. Wm W. Washington, D. C.: The Brookings Institution, 1982. Breheny, Michael and Peter Hall. "Urban Decentralization and Retail Development: Anglo- American Comparison." Wm 13, 4( 1989): 244-261. Brown, Stephen. "Harold Hotelling and the Principle of Minimum Differentiation." Emsmsr‘nflumanfleomphx (1989): 471-493. Burns, Alvin C. and James W. Gentry. "How 'Irnportant' Are Evaluative Criteria rrr Shopping Center Patronage?" mumaigmetafling, 53, 4 (1977-78): 73- 86. Christaller, W. W trans. by C. W. Baskins. Englewood Cliffs, NJ.: Prentice-Hall, 1966. Clements, Donald W. "Utility of Linear Models in Retail Geography." mg Geography, 543, 1 (1978): 17-25. Craig, C. Samuel, Avijit Ghosh, and Sara McLafferty. "Models of the Retail Location Process: A Review." W 60, l (1984): 5-35. Dahl. Robert. W. New Haven: Yale University Press, 1961. Davies. R.L. "Evaluation of Retail Store Attributes and Sales Performance." W Wu. 7, 2 (1973): 89-102. Dawson, Scott, Tom Gillpauick, and Bruce Stern. ”An Empirical Update and Extension of Patronage Behaviors Across the Social Class Hierarchy." Wm Research. 17 (1990): 833- 838. Derbes, Max J., Jr. "Economic Life Concepts." IheAppnjsaLiqumai (April 1987): 216- 224. DiGaetano, Alan and John S. Klemanski. "Urban Regimes rn Comparative Perspective: The Politics of Development rn Britain." WW1): 29,1 (September 1993): 54-83. Doherty, Peter A. "Consumer Research and rts Role rn Shopping Center Development." Adxancesrnficnsumetfieseatch. 18 (1991) 453-461 Fainstein, Susan S. "Rejoinder to: Questions of Abstraction in Studies in the New Urban Politics." W13 (3) (1991): 281-287. Fainstein, Susan S. and Norman 1. Fainstein. W. New York: Longman, Inc., 1983. Feinberg, Richard A., Jennifer Meoli, and Lori Westgate. "A Reinforcement-Affect Model of Mall Patronage?‘ Adxancesrnficnsumetkeseareh 18 (1991) 441-444 Ferbt (. Gall: Garr Gau' Gho th Ght Gh Gt G. 174 Ferber, Robert. "Variations in Retail Sales Between Cities." Wag, (January 1958): 295-303. Galleshaw, Mark. "Market-wide External Obsolescence." WM (October 1991): 519-524. Garreau, Joel. "Edge Cities in Profile." WM, (February 1994): 24-33. . WEE. New York: Doubleday, 1991. Gautschi, David A. "Specification of Patronage Models for Retail Center Choice." Journal Wm XVIII (May 1981): 162- 174 Ghosh, Avijit. "The Value of a Mall and Other Insights from a Revised Central Place Model. " W, 62, 1 (1986): 79-97. Ghosh, Avijit and C. Samuel Craig. "An Approach to Determining Optimal Locations for New Services." WW XXIII (November 1986): 354-362. . "Formulating Retail Location Strategy in a Changing Environment." Was, 47 (Summer 1983): 56-68. Ghosh, Avijit and Charles A. Ingene. eds. MethodLandAnnliszations Greenwich: JAI Press. 1991. Ghosh, Avijit and Sara McLafferty. "Guest Commentary: The Shopping Center: A Restructure of Post-war Retailing. " Wing, 67, 3 (1991): 253- 268. . WW5. Lexington: Lexington Books, 1987. . "Locating; Stores in Uncertain Environments: A Scenario Planning Approach." lonmalLQLRetaiiing, 58, 4 (1982): 5—23. . "Multipurpose Shopping and the Location of Retail Firms." Geographical Annuals, l8, 3 (1986): 215-226. Golden, Linda L. and Mary R. Zimmer. "Relationships Between Affect, Patronage Frequency and Amount of Money Spent with a Comment on Affect Scaling and Measurement." MW 13 (1986): 53- 57 Golledge, Reginald G. and Harry Timmermans. "Applications of Behavioural Research on Spatial Problems I Cognition. " Bmgrcssinflumanfieogranhx. March (1990): 57- 99. Goodchild, Michael F. ":ILACS A Location-Allocation Model for Retail Site Selection." ImmalnLRetailing 60,1(1984):84-100. Granbois, Donald. "An Integrated View of the Store Choice/Patronage Process" AdxancesmfionsumeLRescarch. 8 (1981): 693- 695 Green, Howard L. "Guest Commentary: Retail Sales Forecasting Systems. " Journal of Rclailing. 62. 3 (1986): 227-231. Ha Hz 175 ."Retailing 1n the New Economic Era." in Stemlieb, G. (ed) W .New Brunswick: Rutgers University, Center for Urban Policy Research, 1988. Greenberg, Michael R. and Frank J. Popper. "Finding Treasure in TOADs." Planning, 60, 4 (April 1994): 24-28. Hall, Bruce F. "Neighborhood Differences 1n Retail Food Stores: Income Versus Race and Age of Population. Frommiafieogranhx 59, 3 (1983): 282-295. Hall, Margin?" 101‘" Knapp- and Chfismpher Winsten. Distributioninflrcatflritainand W Oxford: Oxford University Press, 1961. Harper, Lucinda. "Census Bureau Lifts Population Forecast Citing Fertility, Immigration, Longevity." W. (December 4, 1992): B 1. Harvey, David. Expianatiminflmgranmn London: Edward Arnold, 1969. WWW. London: Edward Arnold, 1973. Hartshom, Truman A. W. 2d ed., New York: John Wiley and Sons, 1992. Hickman, Edgar P, James P. Gaines, F. Jerry Ingram. "The Influence of Neighborhood Quality on Residential Property Values " WWW (Summer 1984): 36-42. Holman, Rebecca H. and R. Dale Wilson. "Time Allocation Dimensions of Shopping Behaviour." Adxancesinsonsnmlieseamh 11 (1984) 29- 34 Howell, Roy D. and Jerry D. Rogers. "Research mto Shopping Mall Choice Behavior." MW 8 (1981) 671-676 HoyLHomer. I1‘ 1111.111 .11”..111 id», .u'.’1 ' 1:1 -.: '1.1‘ .1. - 1 . - ~ . New York: National Retail Merchants Association, 1969, Huff, David L. "Defining and Estimating a Trading Area." Waring, 28 (July 1964): 34-38. Ingene, Charles A. "Determining the Distribution of Retail Store Types Within A Metropolitan Area: Macroretail Structure." In W, ed. G. Fisk, 173- 214. New York: Praeger, 1986. . "Structural Determinants of Expenditures on Consumer Services." In B l . M l . . . . . . CT.: JAI Press, 1991. WW . AvijitGhosh and Charles A. Ingene, eds. Greenwich, . "Structural Determinants of Market Potential." W, 60, 1 (1984): 37-64. Ingene, C. and J. Brown. "The Structure of Gasoline Retailing." IonmalnLRetajiing 63 (Winter 1987): 365-92. 176 Ingene, C. and R. Lusch. "Market Selection Decisions for Department Stores." 1mm Retailing. 56.3 (1980): 21-40. Johnston. R.J. WWW New York: John Wiley & Sons, 1980. Kantor, Paul and H. V. Savitch. "Can Politicians Bargain With Business: A Theoretical and Comparative Perspective on Urban Development." ummAffairLQuamdy, 29, 2 (December 1993): 230-255. Kasarda, John D. "People and Jobs on the Move: America' s New Spatial Dynamics" in Stemlieb, G. (ed) City. New Brunswick: Rutgers University, Center for Urban Policy Research, 1988. Kellerman, Aharon. "The Suburbanization of Retail Trade: a U.S. Nationwide View." Qenfnnim 16 (1985): 15-23. . "U.S. Nationwide Trends in Metropolitan Trade: a Perspective of the Mid 19803." Wm, 19, 2 (1988): 179-187. Kern, Clifford, Robert J. Parcells, and Glen E. Weisbrod. "A Disaggregate Model for Predicting Shopping Area Market Attractron W, 60, l (1984): 65- 83. Klaascn, Romain L. "Flckle Obsolescence." W31 (July 1989): 327-337. Kohsaka, Hiroyuki. "A Monitoring and Locational Decision Support System for Retail Activity " Ema'mnrnentandflanninu 25 (1993): 197- 211 . "A Spatial Search-Location Model of Retail Centers." W Analysis. 21, 4 (1989): 338-349. Korgaonkar, Pradeep K. "Consumer Preferences for Catalog Showrooms and Discount Stores: The Moderating Role of Product Risk." WM, 58, 3 (1982): 76- 88. La Londe, Bernard J. "The Logistics of Location. ' Assnciatinn. Chicago: American Marketing Association, (1961), 567-575. Landis, John D. "The Future of America's Central Cities." W 13, 4 (1989): 223-243. Lang, MlchaelH - ' -. Cities. Cambridge: Bailinger Publishing Co., 1982. Langrehr, Frederick W. "Retail Shopping Mall Semiotics and Hedonic Consumption." WM 18 (1991): 428-433 Laulajainen, Risto. WW. Dordrecht, Holland: D. Reidel Publishing, 1987. Law, Christopher 1n association with E. K. Grime, C. J. Grundy, M. L. Senior, and J. N. Tuppen WW London Routlege 1938 177 Leven, C. "Post- industrialism, regional change and the new urban geography,"1 Stemlieb, G. (ed) City. New Brunswick: Rutgers University, Center for Urban Policy Research, 1988. Levy, Frank. "Guest Commentary: Income Trends, the Return to Traditional Values, and the Vanishing Middle Class." mung, 65, 2 (1989): 137-143. Lin, B ..C "The Relationship Between Population, Income, and Retail Sales 1n SMSAs, 1952-1966fW1RmmLEeonomman1Busmcss. 10 (Spring 1970): 25-40. Lloyd, Robert and Diana Jennings. "Shopping Behavior and Income. Comparisons 1n an Urban Environment." Emuomrcfleogranhy. 54. 2 (1978): 157-167. Lord, J. Dennis. "The Malling of the American Landscape," in Dawson, J. and Lord, I. D. (eds.) WW. London: Croon Helm. 1985. . "Retail Decentralisation and CBD Decline 111 American Cities." Working Paper 8802. W University of Stirling, 1988. Mahajan, Vijay, Subhash Shanna, and D. Srinivas. "An Application of Portfolio Analysis for Identifying Attractive Retail Locations." Inumainflietailing, 61, 4 (1985): 19-34. Management Horizons. W. Dublin, Ohio: Management Horizons--A Division of Price Waterhouse, 1990. Mason, Joseph Barry. "Guest Editorial: Redefining Excellence in Retailing." 1.9an Retailing.62, 2 (1986): 115-119. May, Eleanor G. "A Retail Odyssey." Wing, 65, 3 (1989): 356-367. May, Eleanor G. and Malcolm P. McNair. "Department Stores Face Stiff Challenge in Next Decade." W, 53, 3 (1977): 47-58. Mercurio, J. "Store Location Strategies." In Davies, RL. and Rogers, D.S. Eds. Stone LocationandSthssessmentReseamh. New York: Wiley. 1984. Monill, Richard L. "A Note: Continuing Deconoentration Trends in Trade." Gmmhjnd Change, 13 (1982): 46-48. Muller, Thomas. "Regional Malls and Central City Sales: An Overview." In snapping M ed. George Stemlieb and James W. Hughes, 177- 199. Piscataway, New Jersey: Rutgers, 1981. Mulligan, Gordon F. "Consumer Travel Behavior: Extensions of a Multipurpose ShOpping Model. " WM; 19,4 (1987): 364-375. Nakanishi, Masao and Lee G. Cooper. "Parameter Estimation for a Multiplicative Competitive Interaction Model—Least Squares Approach. " Marketing mu, XI (August 1974): 303- 311. 178 Nowakhtar, Susan and Richard Widdows. "Research Note: The Structure of the General Merchandise Industry, 1959-1983: An Empirical Analysis." WW. (1986): 426- 434. Nyden, Philip W. and Wim Wiewel. Eds. W. New Brunswick: Rutgers University Press, 1991. O'Neill, Robert E. "Monitor/NRB Shopping Center Census." Mnnitm, November/December, 1991, 21-44. Pearce, Michael R.and Chow Hou Wee. "Patronage Behaviour Toward Shopping Areas: A Proposed Model Based on Huff‘s Model of Retail Gravitation." Adyanggnin ConsumeLReseareh. 12 (1985): 592- 597 Peterson, Paul E. CityLimits. Chicago: The University of Chicago Press, 1981. Poole, M. A. and P. N. O'Farrel. "The Assumptions of the Linear Regression Model. " Iransaeuenmtthelnstimmtflntishfieegranhers. 52 (1971). 145-158. Public Affairs Counseling. "The Dynamics of Neighborhood Change" Denamnenmf Hmrsingandllrbmflevelenment (May 1975). 1-31. Reilly, W.J. Wayland New York: William J. Reilly, 1931. Rogers, David. WW8. Oxford: Longman Group, 1990. . "Fitting the Format to the Market." Wm September 1984. Rogers, David and Howard Green. "A New Perspective on Forecasting Store Sales: Applying Statistical Models and Techniques 1n the Analog Approach, Cengmphigni Review, 69 (1979): 449-458. Rusk, David. WW3. Baltimore: The Johns Hopkins University Press, 1993. Sheth, Jagdish N. "Emerging Trends for the Retailing Industry." W59, 3 (1983): 6-18. Smith, Geoffrey C. "Shopping Perceptions of the Inner City Elderly." W, 16, 3 (1985): 319-331. Smith, Paul E. W New York: The National Retail Dry Goods Association, 1956. Solomon, Arthur P. and Kerry D. Vandell. "Altemative Perspectives on Neighborhood Decline." IAEA (Winter 1982): 81-98. Stegman, Michael. ”Neighborhood Classification and the Role of the Planner 1n Seriously Distressed Communities." leumalnfltheAmmieanInstitutenLElanners (October 1979): 495-505. Stemlieb, George et. al. "Housing Abandonment 1n the Urban Core." 19mm Amerieanlnstitrrteetfllanners (September 1974): 321- 332 ‘ 179 Stemlieb, George and JamesW. Hughes, ed. Nation.Reelen_and_Metrep.elis New Brunswick Rutgers 1988 New Brunswick: Rutgers University, 1991. Stone, Clarence N. "The Hedgehog, The Fox, and the New Urban Politics: Rejoinder to Kevin R. Cox." MW 13 (3) (1991): 289-297. . Lawrence, Kansas: The University Press of Kansas, 1989. Taylor, M.J. "Space and Time in Industrial Linkage." Area 12 (1980): 150-152. ULI. 11111 ‘ '91"11111c-1011H 01111.15 3.0'5-1111110. .‘u' Washington D. C.: Urban Land Institute, 1977. U.S. Department of Commerce, Bureau of the Census. 1W, Subject Series RC 87-8-3. Washington D.C.: GPO, 1989. . WW GF/914. June 1993. . Qovemmemflnaneesfioi. 1993. . StatisfiealAhstraet. 1993. Varady. David P Nerghborheedllpgradrng—ARealistieAssessment New Yorle State University of New York Press, 1986. Wee, Chow Hou. "Shopping Area Image: Its Factor Analytic Structure and Relationships with Shopping Trips and Expenditure Behavior. AdyancesinCQnsnmeLResemh. 13 (1986): 48- 52. Weinstein, Jerome 1 "Negative Property Values: The Problem of Abandoned Building." IheAuntaisaLIQumal (July 1973): 395- 403 White, Harrison C. "Multipliers, Vacancy Chains, and Filtering 1n Housing. Journalgf theAmeneanImtinlteefElanners (March 1971): 88 94 Wilson, B. L. "Modem Methods of Sales Forecasting." In Davies, R. L. and D. S. Rogers, ed. StemLeeationandStemAssessmentReseareh New York John Wiley & Sous Ltd.,1984. Wind, Y. and V. Mahajan. "Designing Product and Business Portfolios." Hmam W59 (January- February 1981): 155-165. Winsburg, Morton D. "Note: Life Cycle Neighborhood Changes in Chicago Suburbs 1960-80." CmttundChange (Spring 1989): 71-80. Zikmund, William G. "A Taxonomy of Black Shopping Behavior." Wing, 53, 1 (1977): 61-72. REFERENCES Agergard, E., J. Allpass, and RA. Olsen. "The Interaction Between Retailing and the Urban Centre Structure: A Theory of Spiral Movement. " Enyimnmenundflanning, 2 (1970): 55-71. Alexis, Marcus, George W. Haines Jr., and Leonard S. Simon. "The Dynamics of Commercial Structure in Central City Areas." mumalgLMarketing, 35, (April 1971): 10-18. "Maximum Likelihood Estimation of Central-City Food Trading Areas." InurnalhflMarketingReseateh. Di (May 1972). 154-159 Ames, David L. and Edward J. Blakely. "Changin gPlaces: American Planning Policy for the 19905. " loumainfflrhanAflairs, 14, 3gln/4 (1992): 423-446. Amrhein, Carl G. and Daniel A. Griffith. SiatistiealAnalvsisteLGeegranhers. Englewood Cliffs: Prentice Hall, 1991. Applebaum, William. "Methods for Determining Store Trade Areas, Marketing Penetration, and Potential Sales." W11, 111 (May 1966): 127- 141. Babbie, Earl. mum. 5th ed. Belmont CA.: Wadsworth Publishing. 1989. Barmash, Isadore. "Perseverance of Discounters Pays Off." ShoppingCentersInday, (January 1993): 6. Berry, Leonard L. "The Low-Income Marketing System: An Overview." 191mm Retain“, 48, 2 (1972): 44-64. Berry, William D. and Stanley Feldman. MnitinleRegnessiminfimetiee. Newbury Park CA.: Sage Publications, 1985. Black, William C. and Robert A. Westbrook. "A Motivation-Based Shopper Typology." InumalgLRetailing, 61, l (1985): 78-103. Blair, AM. and D.C. Price. WWW. London: Bellhaven Press, 1989. Bowlby, Sophie, Michael Breheny, and David Foot. "Store Location. Problems and MethodS--Ch008iug the Right Sight." RetarliDlstnhutlenManagernent. (January/February 1985): 44-48. ."Store Location: Problems and Methods--Expanding Into New Geographical Areas." Wm (November/December 1984): 41-46 180 181 "Store Location: Problems and Methods--Is Locating a Store Becoming More Difficult?" WumhumManagemmt, (September/October 1984): 31- 33. ."Store Location. Problems and Methods--Local Site and Store Evaluation Issues." Wm (March/April 1985): 4044/ Bowman, John H., Susan MacManus, and John L. Mikesell. "Mobilizing Resources for Public Services: Financing Urban Governments." W, 14, 3/4 (1992): 311-335. Brown, Julia A. N. and Roland T. Rust. "Estimation and Comparison of Market Area Densities." Wing, 62, 4 (1986): 410-430. Brown, Stephen. "An Alternative Paradigm for Retail Planning." Iheflmnen (December 1987): 27-30. . "Institutional Change in Retailing: a geographical interpretation." BrogresanHnmanGeographv, (1986): 181-206. . RetailLoeatiomAMiermSoaleEerspeotive Brookfield USA: Avebury. 1992. . "Retail Location Theory: The Legacy of Harold Hotelling." 1911mm Retailing. 65. 4 (1989): 450-470. . "The Wheel of Retailing." W, 66, 2: 143-147. Christman, Edward. "Can Power Centers Go the Distance." Show. (November 1989): 1. Clark, W.A.V. and PL. Hosking. StatisticalMethodsfioLQeogranhm. New York: John Wiley & Sons, 1986. Cottrell, James L. "An Environmental Model for Performance Measurement in a Chain of Supermarkets." immaioLRetniling, 49, 3 (1973): 51-63. Cox, Eli P. and Leo G. Erickson. Won. East Lansing, MSU Business Studies, 1967. Cronin, J. Joseph Jr. and Steven J. Skinner. "Marketing Outcomes, Financial Conditions, and Retail Profit Performance." humalofiRetaiiing, 60, 4 (Winter 1984): 9-22. Curry, Bruce and Luiz Moutinho. "Computer Models for Site Location Decisions." WW 20 4 (1992): 1247 Davis, Keri and Leigh Sparks. "The Development of Superstore Retailing 1n Great Britain 1960-1986: Results from a New Database." WM Geographers. 14(1989): 7489. Davies, R.L. and D.S. Rogers, ed. W. New York: John Wiley & Sons Ltd., 1984. Dawson, John A. and J. Dennis Lord. ° Prom. New York: Nichols Publishing, 1985. 182 Dent, Borden D. "Metropolitan Retail Structure." In IntenpretEantLCim, ed. T.A. Hartshom, 338-360. New York: John Wiley & Sons Ltd., 1980. Dicken Peter and Peter E Lloyd Loeationin.Space:.Iheoretiea1£erspeotives_in Mammography. 3d ed. New York: Harper Collins Publishers, 1990. Doocey, Paul. "Powering Up the Midwest." Stores, (March 1993): 32-34. Duerksen, Christopher J. and Clifford L. Weaver. "Central Business District Planning and the Control of Outlying Shopping Centers." UrbanLanLAnnnal, 14, 11 (1977): 57- 79. Dunkin, Calvin P. and Rom J. Marking. "The Transformation of Retailing Institutions: Beyond the Wheel of Retailing and Life Cycle Theories." WW, (Spring 1981): 58- 66. Eggers, Mitchell L. and Douglas S. Massey. "The Ecology of Inequality. Minorities and the Concentration of Poverty, 1970-1980. " Americanioumalofifiociology, 95, 5 (March 1990): 1153-88. 1311016011. Rodney A. and Marylynn Gentry. "Suburban Nucleations." W Review. 75 (1985): 19-31. Etgar, Michae. "The Retail Ecology Model: A Comprehensive Model of Retail Change." ResearehirLMarketing 7 (1984): 41- 62 Feinberg, Richard A. and Jennifer Meoli. "A Brief History of the Mall." Anyanmin ConsumeLRrseareh. 18 (1991): 426-427. Finn, Adam. "Characterizing the Attractiveness of Retail Markets." Wing, 63, 2 (Summer 1987), 129-162. Flanigan, James. "Discounters Are Part of Worldwide Trend." WW, 2 May 1992, 14 (C). Fotheringham, AS. and R. Trew. "Chain Image and Store-Choice Modeling. The Effects of Income and Race." EmdmnmentnngifiianningA, 25 (1993): 179- 196. Franz2 Lori and Edward Robb. Thompomielmpactoflflalmsiorrsonflostm O Utv'eU|"l'\.|A-.U.| 0| Ar-u-IUO Staten Columma, Missouri: B&PA Research Center and College of Business and Public Administration, University of Missouri, 1992. Hauser, D. P. "Some Problems 1n the Use of Sthwise Regression Techniques in Geographical Research," CanadianfieogmphenXVIII, 2 (1974):148-158. Gable, Myron, Richard T. Hise, J. Patrick Kelly, and James B. McDonald. "Factors Affecting the Performance of Individual Chain Store Units: An Empirical Analysis." IoumaloLRetailing. 59, 2 (1983): 22-39. Ghosh, Avijit and Charles A. Ingene. "Consumer and Producer Behavior m a Multipurpose Shopping Environment." Wm,22,1 (1990): 70- 93. 183 Gibbs, Anne. "Retail Innovation and Planning." Emgressjnflanning, 27 (1987): 1-67. Goldman, Arieh. "Do Lower-Income Consumers Have a More Restricted ShOpping Space?" .InumaiofiMarketing. 40 (January 1976): 46-54. Goldstucker, J. L, D. N. Bellenger, T. J. Stanley, and R.L. Ottie. Georgia State University, College of Business Administration, 197 8. Green, Howard L. and David S. Rogers. "Changes in Consumer Food Expenditure Patterns." JoumaloLMadteting. (April 1978): 14-19. Greenwood, Michael J. and Richard Stock. "Patterns of Change in the Intrametropolitan Location of Population, Jobs, and Housing: 1950-1980." humming; Economics, 28 (1990): 243-276. Guy, Clifford M. "The Assessment of Retail Impact." IheEiannet, (December 1987): 31- 34. ."Estimating Shopping Centre Turnover: A Review of Survey Methods." 20, 4 (1992):18-23. Hammond, Robert and Patrick McCullagh. WWW Inmdnetion. Oxford: Clarendon Press, 1974. Harris, Frank and Larry O'Brien. WW. London: David Fulton Publishers, 1991. Hazel, Debra. "Making Way for the Next Generation: Tomorrow's Developers Must Be Well-Versed 1n New Types of Disciplines." ChainStomAgeExeoutiye, (October 1988): 33- 38. Hines, M.A. 2d ed. ShoppingSenterDevelopmenLandJnvestmem. New York: John Wiley & Sons, 1988. Hirschman, Elizabeth C. "A Descriptive Theory of Retail Market Structure." Joumaiof Retailing, 54, 4 (1978): 29-49. Hirschman, E. C. and Stampfl, R.W. eds. ' ' ' WChicagm American Marketing Association, 1980. Chicago: AmericanMarketing Association, 1981. Hitt, Jack. "More and More..." WM 23 January 1994, 34- 35. Hollander, Stanley C. "A Rearview Mirror Might Help Us Drive Forward--A Call for More Historical Studies in Retailing." loumaloflRetniling, 62, 1 (1986): 7-10. Hollander, Stanley C. and Glenn S. Omura "Chain Store Developments and Their Political, Strategic, and Social Interdependencies." WM, 65, 3 (1989): 299- 325. 184 Houston, Franklin S. and John Stanton. "Evaluating Retail Trade Areas for Convenience Stores." JonmaLoLRetailing, 60, 1 (1984): 124-136. Howe, Deborah A. and William A. Rabiega. "Beyond Strips and Centers--The Ideal Commercial Form" Ionmalofltheéunerieanflanm’ngAssoeiation, 58, 2 (Spring 1992): 213- 219. Hubbard, Raymond. "Parameter Stability in Cross-Sectional Models of Ethnic Shopping Behaviour. EnyiionmentontiflanningA. 11, (1979): 977-992. _. "A Review of Selected Factors Conditioning Consumer Travel Behavior." W 5 (June 1978): 1-21 Jackson, Donald Dale. "It's Wake-Up Time for Main Street When Wal-Mart Come To Town." Smithsonian, (October 1992): 36-47. Jacober, Steve. "Target Unveils Greatland." DisgountMorchandisoi, (October 1990): 56. James, Franklin J. "Economic Impacts of Private Reinvestment in Older Regional Shopping Centers." GrowthondChango, (1984): 11-24. Jennings, Diana and Robert Lloyd. "Shopping Behavior and Income: Comparisons m an Urban Environment." Economiefioography, 54, 2 (April 197 8) 157- 167. Johnson, Jay. "Alexander's: The City Discounter." DiscountMerehandisor, (February 1991): 20. _. "The Future Role of the Regional Discounter." WW, August 1990): 22. ."Intemational Mass Retail Association: Looking Ahead." Discount m. (May 1990): 26 . "Kmart's Combo Strategy." DisconntMeiohandooi, (December 1992): 36- 40. Kaikati, Jack G. "Don't Discount Off-Price Retailers." HMBusjnessReyiew, (May- June 1985): 85-92. Kering, Roger A., Vijay Mahajan, and Subhash Shanna "Assessing Market Penetration Opportunities and Saturation Potential for Multi-store, Multi-market Retailers." loumal W. 64. 3 (1988): 315-333. Knee, Derek and David Walters. ' ' ° -- Southampton: The Camelot Press, 1985. Korgaonkar, Pradeep K. "Shopping Orientations, Importance of Store Attributes, Demographics and Store Patronage: A Multivariate Investigation. " Ahonmsinmano W, (Winter, 1981): 34-38. Lesser, Jack A. and Marie Adele Hughes. "Towards a Typology of Shoppers." Businoss Horizons, (November/December 1986): 56-62. 185 Levy, Walter K. "Are Department Stores Doomed?" DisoopntMorohanoisor, (May 1991): 56. . "The End of an Era: A Time for Retail Perestroika." Journaloflfiotailing, 654, 3 (1989): 389-395. Lewis, Gerald. "Strategic Management Concepts in U. S. Food Retailing and Wholesaling " IntemationalezumaloWhufionManagement 19.6 (1991): 21- 26. Lilien, Gary L. and Ambar G. Rao. "Emerging Approaches to Retail Outlet Management " SloanManagemmtReldew. (Winter 1979): 27-36. Lord, J. Dennis and Charles D. Lynds. "The Use of Regression Models 1n Store Location Research: A Review and Case Study." AkronBrlsinessandEonomieReview. (Summer 1981): 13-19. MacKay, David B. and Richard W. Olshavsky. "Cognitive Maps of Retail Locations: An Investigation of Some Basic Issues." LoiimalofConsiimeLResoaroh, 2 (December 1975): 197-205. Madden, C. S. andJ. V. McNeal. "Retail Failure. A Site Selection Perspective." Southwestlournalofflusinessandficonomies. 4 3(1987): 18- 28. Mahajan, Vijay, Subhash Sharma, and Roger A. Kerin. "Assessing Market Penetration Opportunities and Saturation Potential for Multi-store, Multi-market Retailers." Journal 9139411111113. 643 (1988): 315-333. Marcus, M. "A Behavioural Approach to Retail Blight." Enyironmontandflanningfi, 15 (1983): 739-750. Mayer, Morris L. "1949-1989: Retail Reflections." IournaloflRotailing, 65,3 (1989): 396-401. Mier, Robert and Wim Wiewel. "Analyzing Neighborhood Retail Opportunities: A Guide for Canying Out A Preliminary Market Study." WM. 198 1. Miles Mike E. et. al ReaLEstateDevelopmmtzhineiplesandfirocess. Washington D.C.: Urban Land Institute, 1991. Morgan, Judy. "Planning for Strong Retail Strips." Qnerationaifloogranhy, 7, 2 (1989): 19-21. Palumbo, George, Seymour Sacks, and Michael Wasylenko. "Population Decentralization Within Metropolitan Areas: 1970-1980." loumaiofllrhanflonomios, 27 (1990): 151-167. Palma, Dolores. "The Top 10 Ways to Revitalize Your Downtown." American City 8; County, (November 1992): 48-57. Pearson, Thomas D. "Location! Location! Location! What is Location?" IhoAnnraisal Journal, (January 1991): 7-20. 186 Peiser,RichardB. t'nu'. 01 {‘r .m‘h '11111I1‘ J 1' 01‘: 1‘ Washington D. C.: Urban Land Institute, 1992. Peterson, Eric. "Power Centers! Now!" Stores, (March 1989): 61-66. Potter, Robert B. "The Multivariate Functional Structure of the Urban Retailing System: a BritishCase Study." IransaetionanstihrteoLBritishfloographers. 6(1981):188- 213. . WWW. Aldershot: Gower Publishing, 1982. Rains, Patrick. "Flight or Fight: The Town That Took On Wal-Mart--and Won." WW (November 1992): 50. Reynolds, Mike. "Who's the Power in Power Centers?" ShoppingCentai-Jflorid, (March 1993): 32-35. Riordan, John T. "How Shopping Centers and Malls Are Changing." Discount Merchandiser. (May 1990): 37- Robertson, Kent A. "Downtown Retail Activity in Large American Cities 1954-1977." GeographieaLRev'rew. 73 (1983): 314-323 Roca, Ruben A., ed. MadratfiesoarehjorShonpingCenters. New York: lntemational Council of Shopping Centers, 1980. Rogers, David. "America's Shopping Centres--A Mid-Life Crisis." Ram] and DiSltibmion MW (November/December 1987): 21-25. ”A Review of Sales Forecasting Models Most Commonly Applied in Site Evaluation." IntemationallournalofRetaiLflrflsnihmionManagemenLZO. 4 (1992): 3-11. Rowley, Gwyn. "British Superstore Retailing Database: A Rejoinder." Iransaotions, InstihlteoLBritishGeographrrs. 15 (1990): 240-244. Russell, Vera Kilduff. "The Relationship Between Income and Retail Sales in Local Areas." IholoumaloflMarkoting, (January 1957): 329-332. Savitt, Ronald. "Looking Back To See Ahead: Writing the History of American Retailing" W 65,3 (1989): 326-355. Shaw, Gareth and Dennis Wheeler. StatisticalleehniouesinfleographioalAnalvsis. New York: John Wiley & Sons, 1985. Schwanke, Dean. "Navigating the Value Retail Marketplace." UrbanLand, (May 1993): 38-42. Sirnkin, Lyndon P. "Evaluating a Store Location." Intemanonalloumalofketatlfi DistributionManagement. 18 ,:4 33- 38. Smith, Paul E. "The Spatial Shopping Behaviour of the Urban Elderly: a Review of the Literature." 9mm. 19, 2 (1988): 189-200. 187 Solomon, Barbara. "Mass Market Report--Power Centers: Fast and Focused." Woman: Many, (November 13, 1992): 14-18. "Mass Market Report-Super Stores: Spreading Out." MM, (June 2, 1993): 20. Somogyi, Jan de. "Retail Planning for the Next Ten Years." Retail & Di strihntj on MW (September/October 1986): 9-13. SPSS IncJMarija J. Norusis. W. Chicago: SPSS Inc., 1990. Stampfl, Ronald W. "The Consumer Life Cycle." IholournalofConsumorAffairs, 12, 2 (Winter 1978): 209-219. Stemlieb, George and James W. Hughes. eds. Shopping Centers: U.S.A. Piscataway: Rutgers, 1981. Thompson, John S. Site_Soleotion. New York: Chain Store Publishing, 1986. Thorpe, David. "The Development of British Superstore Retailing--Further Comments on Davies and Sparks." W 16 (1991) 354- 367. Van Tassel. Charles E. AnalysisofliaeiorslnflueneingRetaiLSales. East Lansing. Marketing and Transportation Paper, MSU Graduate School of Business Administration, 1966. Vastine, William J. "The Convenience Store Industry 1n the 19905 and Beyond. " mmanonaLIoumalofRemnmsmhunonManagement 19.6 (1991): 27-31. Walker, Gail. "Developers Put the 'Pow' 1n Power Centers With Value, Convenience, and Critical Mass." MalueRetailNews (May 1992): 4-6. Wechsler, Pat. "City Business: Attention Discount Shoppers. . ." BusinossNY, (October 1992): 3 1-33. Westbrook, Robert A. ”Sources of Consumer Satisfaction with Retail Outlets." Journal of Retailing. 57, 3 (1981): 68-85. Whysall, Paul "Changing Planning Policies for Large Stores." Botai] 5; Q1“ Snibtltion Management. (January/February 1985): 8-12. W— "The Power Retailer5--Special Report." (August 1992): 2-15. Wrigley. Neil. ed. StoreoholeerstoreJoeationandmarketanalvsis London and New York: Routledge, 1988. Zelner, Wendy and Marti Benedetti. "Clout! More and More, Retail Giants Rule the Marketplace." Businessfleok, (21 December 1992): 66-72. nrcuran STRTE UNIV. LIBRARIES 111111111111111111111 "111111 ”1 11111111111111 31293014214971