‘ mm; «m ii iii] Miami Will ” 3 1293 01103 0099 -i i i - -umvx. 6 This is to certify that the dissertation entitled THE RELATIONSHIP BETWEEN DOMESTIC DEMAND AND U.S. EXPORTS: A TEST OF THE DEMAND PRESSURE HYPOTHESIS presented by Michael Raymond Myler has been accepted towards fulfillment of the requirements for Ph . D. . Economics degree m ( L/(Yé/[C [L5H I(L{',, :1, L L/ Major professor Dateff ID- 72 72' /LH/B MSUis an Affirmatiw Action/Equal Opportunity Institution 0-12771 MSU LIBRARIES m RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. ,. JUN: 1. 8 199i THE RELATIONSHIP BETWEEN DOMESTIC DEMAND AND U.S. EXPORTS: A TEST OF THE DEMAND PRESSURE HYPOTHESIS BY Michael Raymond Myler A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Economics 1983 4/ ‘U I ‘v’ ABSTRACT THE RELATIONSHIP BETWEEN DOMESTIC DEMAND AND U.S. EXPORTS: A TEST OF THE DEMAND PRESSURE HYPOTHESIS By Michael Raymond Myler An inverse relationship between domestic demand and eXport performance has been hypothesized by several writers. Empirical tests of this prOposition (called the Demand Pressure Hypothesis) have, to date, yielded mixed results. The usual test has involved a single-equation model with the eXport quantity as the dependent variable and the eXport price, some measure of world demand, and an indicator of domestic demand pressure as the eXplanatory variables. This dissertation develOps a structural model of supply and demand at the commodity level. From this simultaneous equation system, reduced forms are derived for eXport price and eXport Quantity. Besides the use of both price and quantity as endogenous variables, this study improves on the literature by the inclusion of factor prices in the supply function. Four models of eXport behavior are tested on U.S. quarterly data for the period 1965.1 through 1979.”. The Michael Raymond Myler tests are run on thirty-one 7-digit commodities from Standard International Trade Classification (SITC) Sections 5, 6, 7, and 8. The four models differ from each other with respect to the effect of the capacity utilization rate on the relationship between domestic demand and eXports (both price and quantity). In the models that depend upon a distinction between low and high demand pressure, four different capacity utilization rates (83, 85, 86, and 87 percent) are tested as the separation rate between low and high pressure for each commodity. The data for twenty-seven of the thirty-one commodities provide at least some support for the Demand Pressure Hypothesis. The Hypothesis is supported in 89 out of 305 tests. For fifteen of the commodities there is evidence that suppliers treat eXports as a residual. An interesting implication of the study is that for several commodities, average total cost curves have horizontal segments. Copyright by MICHAEL RAYMOND MYLER 1983 In memory of my mother, Emily G. (Yetke) Mokszycke. iii ACKNOWLEDGMENTS This research project has taken a long time to complete, and over the years the debts have accumulated. My initial interest in economics can be traced to Howard Swaine of Northern Michigan University. Alan Nichols and Habib Zuberi of Central Michigan University encouraged me to pursue a doctoral program. My thanks to all three. Albion College was most generous in making its computer a free good (in the language of the public choice theorist, the individual consumption-payment link was broken). John Williams, Director of Academic Computing at Albion, spent many hours loading the econometric software onto the Burroughs 6700 and debugging it; and he freely furnished programming advice, even during his vacations. I owe a debt of gratitude to James McCarley, Professor of Economics at Albion College, for a key methodological insight. Arthur Mullier of the United Nations furnished data on world industrial production. The contributions from two Albion students deserve special recognition. Donald Luciani patiently and accurately tabulated the preliminary statistical results and devoted a Christmas break to the summarizing of several articles. In the final moments of the dissertation, deadlines could not have been met if Scott Harrison had not willingly set aside his other iv res hel 58 'v' reSponsibilities and helped with the editing of the text. The members of my dissertation committee were most helpful and deserve special thanks. Ed Sheehey offered several suggestions that encouraged me to be more explicit about the conclusions to be drawn from the econometric analysis. Lawrence Officer's questions and comments are reflected at several points in the manuscript. Mordechai Kreinin served as chairman of the committee. His careful scrutiny of the drafts and detailed suggestions for revisions contributed crucially to the scholarly content of the final document. Without his advice and encouragement, this dissertation could not have gotten past the proposal stage. H s-c +4 ’9? d-‘ TABLE OF CONTENTS List of Tables ... ........... ..... ................... vii1 List of Figures ..................................... x Chapter 1: Introduction ........................ .... 1 Chapter 2: Review of the Literature .. .............. 5 I. Introduction ............................... 5 II. The Relationship between Exports and Domestic Demand: Theoretical Development ....... 10 A. Macroeconomic Theory ..... ........ ... 11 B. Microeconomic Theories .............. 12 C. The Current Status of the Debate .... 25 III. Empirical Development of the Demand Pressure Hypothesis ..... ............... 30 A. Three Estimates of Export Functions ........................ 30 B. Tests of the Demand Pressure HypotheSiS 0....0..00.00 00000000 .0 32 IV. summary 000.00.000.00...000..00000000....000 “0 Chapter 3: The Theoretical Basis for the Demand Pressure Hypothesis . ........... . ..... ...... A1 I. Derivation of the Demand-for-Exports FunCtion 0.00.0...0...00.000..0 00000000 L11 II. Derivation of the Export Supply Function..... AA A. The Domestic Demand Function......... “4 B. The (Total) Supply Function ......... uu C. The Export Supply Function ..... ..... 59 III. The Complete Models ..... ............. .. ..... 63 Chapter 1‘: Data 0....0000.000000.0000...00..........000 68 I0 IntrOdUCtion 000.0 000000 ..000000000.00.000..0 68 II. Trade Data ....... ........... ...... .......... 68 III. Macroeconomic Data .. ..... ................... 72 A. world output .0.........00....0...0.0 72 B. Dollar Exchange Rate ....... ...... ... 73 vi C. U.S. Output .......... . .............. 7“ D. Demand Pressure ..... .... ........... . 75 E. Input Prices ............. ......... .. 75 Chapter 5: Results of Regression Analysis ...... . ....... 77 I. Estimation Procedure ...................... .. 77 II. Discussion of Results ... ...... .. ............ 82 A. Overview ........... . .... ....... .. 82 B. Commodities from SITC Section 5 (Chemicals) .... . .............. 87 C. Commodities from SITC Section 6 (Manufactures) ...... . ............. 96 D. Commodities from SITC Section 7 (Machinery) . .................... 100 E. Commodities from SITC Section 8 (Miscellaneous Manufactures) ..... 106 F. The Tables .......................... 109 Chapter 6: Conclusion .................................. 111 I. Summary ............... . ......... . ..... 111 II. Comparative Performance of the Models ...... 115 III. Comparative Performance of Cut- Off Rates ... 116 IV. Henry's Weak Test .......................... 121 V. Exports As 3 Residual .. .................... 122 VI. Evaluation of DPH ..... . .................... 12M VII. Future Research ........ . ................... 125 Appendix 1: Calculation of XRIMF ....................... 128 Appendix 2: Calculation of PL and PK ................... 132 Appendix 3: Tables of Regression Results ..... .... ...... 135 Bibliography .000000.0000 ..... 00.000.000.00 000000000 000.0215 vii LIST OF TABLES BEBE 5- 1 Summary of Regression Results ............. ..... . 85 A- 1 Iron oxides and hydroxides, pigment grade ........ 135 A- 2 Iron oxides and hydroxides, pigment grade ........ 136 A- 3 Iron oxides and hydroxides, pigment grade ........ 137 A- A Iron oxides and hydroxides, pigment grade ........ 138 A- 5 Printing inks ...... ........ ...... ...... .......... 139 A- 6 Printing inks ........................... ...... ... 1A0 A- 7 Rubber cement . ...... ....... ................... ... 1M1 A- 8 Rubber cement ..... ........ .... ....... ............ 142 A- 9 Newsprint paper .................. ... ............. 1A3 A-10 Newsprint paper ........... ....................... 1AA A-11 Newsprint paper ...... ................. . .......... 1A5 A-12 Newsprint paper ...... ......... ..... . ............ . 1A6 A-13 Pig iron, including cast iron .................... 1A7 A-1u Pig iron, including cast iron .................... 1A8 A-15 Pig iron, including cast iron .................... 1A9 A-16 Concrete reinforcing bars .. ...... .. ......... ..... 150 A-17 Concrete reinforcing bars ............. ........... 151 A-18 Copper alloy wire, bare ....... ................... 152 A-19 Copper alloy wire, bare ................ ..... ..... 153 A-20 Copper alloy wire, bare .......................... 15A A-21 Copper and copper alloy powder and flakes ........ 155 A-22 Copper and copper alloy powder and flakes ........ 156 A-23 Aluminum and aluminum alloy wire, not insulated... 157 A-2A Aluminum and aluminum alloy wire, not insulated... 158 A-25 Aluminum and aluminum alloy powder and flakes .... 159 A-26 Aluminum and aluminum alloy powder and flakes .... 160 A-27 Zinc and zinc alloy sheets, plates, and strip .... 161 A-28 Zinc and zinc alloy sheets, plates, and strip .... 162 A-29 Door and window sash, frames, moulding, and trim of iron and steel ........................ 163 A-3O Door and window sash, frames, moulding, and trim of iron and steel ........ ..... ...... ..... 16A A-31 Door and window sash, frames, moulding, and trim of aluminum .............................. 165 A-32 Door and window sash, frames, moulding, and trim of aluminum .............................. 166 A-33 Hacksaw blades, hand and power ................... 167 A-3A Hacksaw blades, hand and power ..... .............. 168 A-35 Twist drills, metal-cutting ...................... 169 A-36 Twist drills, metal-cutting ...................... 170 A-37 Safety-razor blades .............................. 171 A-38 Safety-razor blades .............................. 172 viii 33 ..fi :4 r». allnflv AH; AHV 1|. Caiaz.r-J pkaCJCJrhvé . . . . . . . - .n - 2 Fe «(4h ‘U s, >>>T>>>> arq-cavq-aarq OOCDN O‘U1 1:1.» Motors, AC, Motors, AC, Motors, AC, not not not not polyphase--induction, polyphase--induction, polyphase--induction, Motors, AC, polyphase--induction, Combines, self-propelled ..... Combines, self-propelled ......... Combines, self-propelled Dozers, for mounting on tractors ............. for mounting on tractors ................. Dozers, Needles, sewing machine Needles, sewing machine .................. Centrifugal pumps for liquids, single-suction, close-coupled, Centrifugal pumps for liquids, single-suction, close-coupled, Centrifugal pumps for liquids, single-suction, close-coupled, Air compressors, stationary, over 100 hp Air compressors, stationary, over 100 hp Air compressors, stationary, over 100 hp Typewriters, standard non-portable, Typewriters, standard non-portable, Radios, household type, without phonograph Radios, household type, without phonograph Shavers, electric ................... Shavers, electric .............. Vacuum cleaners, electric, household type Vacuum cleaners, electric, household type Vacuum cleaners, electric, household type Vacuum cleaners, electric, household type Toasters, automatic, electric, Toasters, automatic, electric, Sun or glare glasses, and sun goggles Sun or glare glasses, and sun goggles Sun or glare glasses, and sun goggles Sun or glare glasses, and sun goggles Clocks, electric Clocks, electric ....................... Clocks, electric ........ Tape, pressure-sensitive, Tape, pressure-sensitive, Pens, ball-point type Pens, ball-point type Pens, Pens, plastic plastic ball-point type .0... 000000 000.... ix ball-point type .............. ..... .. single-stage, under 2-inch ... single-stage, under 2-inch single-stage, under 2-inch ... electric, electric, over 20 hp.. over 20 hp.. over 20 hp.. over 20 hp.. household type household type Bags 173 174 175 176 177 178 179 180 181 182 183 18A 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 20A 205 206 207 208 209 210 211 212 213 21” 67‘9U1l 121:2 . u . u u o - ~ 3.43333“ CJEJOIO LIST OF FIGURES 2-1 The Dumping Model: Infinitely Elastic Demand for Exports .............. ...... ........ ..... 1U 2-2 The Dumping Model: Less Than Infinitely Elastic Demand for Exports ........................... 15 2-3 The Dumping Model: Horizontal Marginal Cost .... 16 Explanation of Variable Names ...................11O Comparison of Capacity Utilization Rates ........120 The Exports-Are-A-Residual Hypothesis ........... 127 2-u The Dumping Model: Downward Sloping Marginal Cost ........ .............. ..... . .... 17 2-5 The Competitive Model: Less Than Infinitely Elastic Demand for Exports ............ ....... 19 3-1 Alternative Definitions of Capacity ....... ..... . 47 3-2 The Effect of an Increase in Domestic Demand on Export Supply ....... ...... ........ ...... A9 3-3 Profit Maximization vs. Capacity Output: Perfect Competition ........................ 51 3-u Profit Maximization vs. Capacity Output: Monopoly ............. ...... .......... ...... 52 3-5 The Effect of an Increase in Domestic Demand on the Quantity Of Exports: The Price- Discriminating Monopolist ......... .......... . 56 3-6 Simultaneous Equation Models .............. ..... . 65 3-7 Reduced Forms ................................... 66 3-8 Definitions of Variables ........................ 67 A-l Description of Commodities Tested .. ............. 70 5-1 Reduced Forms To Be Estimated ........ .. ........ 81 5 2 6 1 6 2 FE *b n\v 3 .flu CHAPTER 1 INTRODUCTION The purpose of this dissertation is to investigate the relationship between domestic demand and a country's eXports. The country selected for the test is the United States. The hypothesis to be tested is that an increase in domestic demand pressure affects exports adversely; this effect can surface as a reduction in the quantity of eXports or as an increase in the price of the eXports. This hypothesis, which shall be called the Demand Pressure Hypothesis (DPH), is of theoretical and practical (or policy-making) interest. As shall be shown in Chapter 3, it appears that what the DPH is all about is the shape (more precisely, the specification) of the export supply function, which in turn depends crucially upon the Marginal Cost function. Planners and analysts engaged in macroeconomic policy will find the results of this project interesting because of the implications the hypothesis has for the eXport function in the standard neo-Keynesian paradigm. In the usual Keynesian-cross presentation, exports--if treated at all--are assumed to be independent of income; the DPH, however, suggests that the eXport function should be plotted with a negative slope. Anti-recessionary demand-management policies, then, cause a deterioration in the balance of Syst trade (or place downward pressure on the exchange value of the currency) not only because imports rise but also because eXports decline. The important result is that part of the increase in domestic demand can be satisfied by a reduction in exports rather than by an increase in domestic production. All other things remaining the same, the foreign purchasers will switch their demands to their own domestic (or other foreign) production. In an analogous fashion, when domestic demand declines, domestic production need not decline by the same amount because resources can be switched from production for the home market to production for the eXport market. The original demand-management policy is thus transmitted to other countries; and, from the point of view of the domestic labor force, the policy's impact on the unemployment rate is weakened. Furthermore, the Hypothesis applies not only to policy-induced changes in domestic demand but also to changes that occur as a routine result of the business cycle or as a result Of shocks to the system. If the evidence is overwhelmingly in favor of the Demand Pressure Hypothesis, then the lessons to be learned by studying the closed-economy macro model are somewhat reduced. The study proceeds along the following lines. Chapter 2 reviews the literature, beginning with the seminal 1965 article by Brechling & Wolfe.1 Chapter 3 is the "theory" chapter; it develops several models that would appear to capture the flavor of the Demand Pressure Hypothesis and suggest the kinds of relationships that would have to show up in the evidence in order to refute or support the hypothesis. A desirable feature of the models is that they are natural outgrowths of basic neoclassical views of firm behavior and thus are well-grounded in microeconomic theory. Chapter A describes the data to be used for testing the hypothesis. Compared with other studies, one unique feature of this study is that the export data used are for seven- digit industries. The greatest degree of disaggregation used in other contributions to this tOpic is that of a four- digit industry. Another unique feature is that the eXport quantity is measured in actual physical units such as tons or board-feet. The commodities chosen for this study come from Standard International Trade Classification (SITC) Sections 5, 6, 7, and 8, which are, respectively, chemicals and related products, manufactured goods classified chiefly by material, machinery and transport equipment, and miscellaneous manufactured articles. Chapter 5 takes the reduced form equations derived in Chapter 3, explains how they were estimated, and then presents the results of the regression analysis. The 1Dunlevy (1980), however, claims that the concept is implicit in the formal models of Nurske (1956). discussion is organized by SITC Section, but on several occasions the similarities or differences between commodities from different sections were important enough to merit explicit mention. Finally, Chapter 6 summarizes the results and suggests some policy implications of these results. CHAPTER 2 REVIEW OF THE LITERATURE This chapter reviews the literature on the relationship between eXports and changes in domestic demand. Section I introduces several versions of the Demand Pressure Hypothesis that can be culled from the literature. Section II traces the intellectual deveIOpment of the Demand Pressure Hypothesis and evaluates the theoretical contributions of several writers. Section III reviews the empirical literature. I. Introduction. The most general statement of the Demand Pressure Hypothesis is that changes in domestic demand have an inverse effect on that country's exports. Phrased in this manner, the Hypothesis encompasses all the variations on the same general theme that have appeared in the literature. The following versions of 3 Demand Pressure Hypothesis can be identified from this literature: 1. Changes in domestic demand pressure lead to changes in the Opposite direction in the quantity of exports. 2. When domestic demand pressure is high, changes in domestic demand pressure lead to Opposite changes in the quantity of eXports. 3. The negative effect on the quantity of eXports as domestic demand pressure increases is greater than the positive effect on this quantity as domestic The demand pressure decreases. This is sometimes called the ratchet effect. A. There is an inverse relationship between the quantity of exports and the rate of change Of domestic demand pressure. 5. Changes in the demand for eXports will have an effect on eXports when domestic demand pressure is low but not when domestic demand pressure is high. 6. The change in the quantity of exports that results from a change in domestic demand can be separated into two distinct changes. First, the change in domestic demand may cause the eXport price to change, which in turn will cause a change in the quantity demanded. Second, the change in domestic demand will cause a change in the quantity of exports in addition to (and independent Of) any price-induced change in the quantity demanded. 7. An increase in domestic demand will cause the equilibrium eXport price to rise and equilibrium eXport quantity to decline. A decrease in domestic demand will cause the Opposite responses. Not all these versions are mutually exclusive. Indeed, the separation between price and non-price effects described in NO. 6 can be made a part Of most of the other versions. The tendency to identify the Demand Pressure Hypothesis as a quantity effect and to ignore the effect Of domestic demand on eXport price led unfortunately to two theoretical problems. First, some writers ignored price completely, as though price were not affected by changes in eXport supply. Second, when it was recognized that price could indeed change, the claim was made that this price change would by itself cause a change in the quantity of exports demanded. This claim is reasonable as long as one keeps in mind that it was the change in the quantity Of eXports producers were willing to Offer at each possible price (that is, a shift in the eXport supply function) that initially caused the change in price; and that as the price now rises, not only will the quantity demanded decline but also the quantity supplied will increase. In pursuit of the effect of domestic demand on eXport quantity, this simultaneous determination of price and quantity was ignored, and it was asserted that the effect on quantity was greater than the change in quantity that arose from a movement along the demand curve. This latter quantity was attributed to changes in relative prices and was called a price effect (it is interesting to note that it was always the movement along a demand curve, never the movement up or down a supply curve, that was called a price effect). The remaining quantity change became known as the non-price effect, the quantity effect, or the capacity effect. In graphical terms, the implication is that the new quantity is not indicated by the intersection of the demand curve and the new supply curve. There is no evidence that the early writers were aware of their excursion into disequilibrium analysis; but later writers, intent on explaining the non-price effect as though it were the essence of the Demand Pressure Hypothesis, recognized that neoclassical economic analysis did not predict such an effect. Scenarios deveIOped to eXplain why ‘h such an effect might exist used the concepts of non-clearing markets and non-price rationing. Whereas the Hypothesis evolved into a claim that the Observed "total" change in quantity was greater than the price-induced change (apparently excluding the possibility of a quantity-induced change in price), more traditional microeconomics would claim that the observed change is only part of the "total" change, where "total" change in this case refers to the change in eXport quantity that would result from a change in domestic demand if the eXport price were to remain constant. This is measured by the horizontal shift in the eXport supply function and is what one might be willing tO call a non-price or capacity effect. The actual--that is, Observed--change in quantity is less than this because the price will change; and a price change will induce a change in the quantity supplied which will Offset part Of the capacity effect. In this view, the observed change is eXplained entirely by what the other view calls the price effect, but this view recognizes that price and quantity are both dependent variables. Recent attempts to give the Hypothesis a firmer grounding in microeconomic theories of producer behavior (while continuing to exclude a relative price effect) have required excursions into non-clearing markets and non-price rationing. Whereas the early papers could simply claim that in: TGC the rle: T101 individual firms viewed exports as a residual, the more recent ones try to give some economic meaning to the concept of "residual," to show why a firm would voluntarily and repeatedly produce excess output that could then be sold in foreign markets. The argument has generally been phrased in terms of a domestic producer who desires to keep his plant running at full capacity regardless of the state of domestic demand for his product--if demand is strong he sells all his output at home, and as demand weakens, he sells increasing amounts of it on the eXport market. Despite the initial intuitive appeal of the eXplanation, deficiencies in the analysis become apparent from the papers that try to build a model of the firm that retains profit-maximization as an Objective, employs the tools of marginalism, and yields the desired conclusions. These deficiencies are that the traditional argument requires a specific definition of capacity (a physical limit to production) and the assumption that price is irrelevant to the producer; that is, quantity supplied is an exogenous variable because the producer keeps the plant running at full capacity. An additional and yet necessary assertion that producers favor the home market has not been adequately justified. The Hypothesis that is deveIOped in Chapter 3 avoids the difficulties just mentioned. The specific wording of the Hypothesis is that increases (decreases) in domestic de EX 1O demand will lead to decreases (increases) in the quantity of exports and to increases (decreases) in the price of exports. This differs from much Of the literature in that there is no artificial distinction between price effects and capacity effects. The desirability of making this departure was becoming evident by the time of Henry (1970) but was not finally asserted and defended until Dunlevy (1980). The Demand Pressure Hypothesis has deveIOped from a macro to a micro phenomenon and from a micro theory relying (at first, unwittingly) on non-clearing markets to a neoclassical microeconomic theory that uses the market- clearing mechanisms implicit in a simultaneous system Of equations. This deveIOpment is traced in Section II below. II. The Relationship between EXports and Domestic Demand: Theoretical DevelOpment. The theoretical treatment of the relationship between eXports and domestic demand can be based on macroeconomic theory or on microeconomic theory. The macroeconomic approach uses the standard Keynesian paradigm. Microeconomic approaches are much more prevalent in the literature; and, in order to analyze these, we shall group them into the following four categories: (1) standard market-clearing models, (2) non-market-clearing models, (3) models that interpret "price" more broadly than usual, so 11 that it includes delivery lags and credit terms, and (A) models that define the commodity more broadly than usual, so that, for example, the availability of repair facilities (or other services) is an integral part of the item being purchased. Subsection A below looks at the Keynesian implications, and Subsection B considers the microeconomic models. A. Macroeconomic Theory. In Keynesian analysis, changes in real national income are the key variable affecting the trade balance and the balance of payments. Because of an implicit assumption that prices are independent of real income changes, eXports are not directly affected by domestic demand. Artus (1970, p. 2A9) points out that in a Keynesian framework eXports and domestic demand are positively related. An increase in national income in an Open economy (Country A) will lead (through the marginal propensity to import) to an increase in that country's imports. This increase in the Rest-Of- the-World's eXports is an increase in aggregate demand, and the resulting rise in foreign national income will lead to an increase in ROW imports from Country A. Thus, in Country A the initial increase in domestic demand caused eXports to increase (through foreign repercussions) rather than to decrease as the DPH predicts. 12 B. Microeconomic Theories. This Subsection discusses the four types of microeconomic explanations of the relationship between eXports and domestic demand. It considers, in turn, market- clearing, non-market-clearing, true-price, and true-quantity models. (1) Market-Clearing Models. There are two models which use the traditional concepts of clearing markets--that of a price-discriminating monOpOlist and that of a competitive industry both at home and abroad. In the international trade literature a price- discriminating monOpOlist model is more often called the dumping model, because it is used to eXplain the behavior of a firm that practices persistent dumping. (For a deveIOpment Of this model in the context of dumping, see Kreinin, 1979, pp. 337-341.) As COOper, Hartley, & Harvey (1970, pp. 52-56), Artus (1970), and Ball (1961) have Shown, as long as the marginal cost curve is upward SIOping, a change in domestic demand will lead to an inverse change in the quantity Of eXports. If foreign demand is infinitely elastic (as in Figure 2-1) the eXport price will not change; but if the firm does have some monOpoly power in the eXport market (see Figure 2-2), then the export price and export quantity will be inversely related. With a horizontal marginal cost curve (Figure 2-3), eXports will not be affected by changes in domestic demand. If the marginal 13 cost curve is downward leping (Figure 2-4), however, an increase (decrease) in domestic demand will lead to an increase (decrease) in the quantity of eXports also. ~00C«£€LCC ADV acute: “Coax: 30 m. UMP/Tut: maid: Auwv W 14 .muwonvfl wow magma owummam .AHOOHGCCH “Hood: wdE—Hd men. .Hnm pupa: man 0 a be Be as as we Ea A .1 .1 w A _ _ _ - _ _ _ _ _ _ . _ _ _ _ _ u _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ . Np _ _ _ _ . m: E. U _ _ x& I .. 1 I I I ... Hm Na 0: BEA-Bo 3 » Hana: 825 So .0. Danna 962 3 w 15 .muwoaxu wow cam-5Q owummau joficch 53:. mmmq ”flown: v.59:— OFF No He xv HmU Nxe 2e Nae Haw . . u ... m NMZ . . . . . _ _ . . . . . . _ . _ . . H . . . . . _ . u . . _ _ ".-----I: u u . . o I u u - 1 u . ...11 33 NE DZ a Seuss £er 3v .0, DER: 25: 3 35980 g .~-N mama 16 N O’ H C’ .umoo HmCAwwm: HmOCONHwo: xv HxU ”specs wfidfia «5. TN wanna go 02 SEA .DG 4 —----—-----df -- -— ---- --1 BEES 6v w umxwmz Duos 3v w umxwm: m8: AS l7 .umoo Hmfiwwmz wfiaon Mommas-e58 O’ NLU d? -------H BEES 3V m 69in: flog 3V busing m8: 3 ”ER: wfiafia ma. TN Pan: 3 18 Consider now the case Of a competitive market. The home country's eXport supply curve and the Rest-of-the- World's import demand curve (that is, ROW's demand for the home country's eXports) can be derived from the appropriate domestic demand and supply curves (for such a derivation, see Kreinin, 1979, pp. 276-279). A change in domestic demand will be associated with an inverse change in the quantity of exports (see Figure 2-5). In addition, if the world import demand function is downward sloping, then the eXport price will change also. If, however, the home country faces an infinitely elastic demand for its eXports, the price of those exports would not change. Both the dumping model and the competitive model yielded the same conclusion. Provided that marginal cost curves are upward SIOping, an increase in domestic demand will reduce eXports and a decrease in domestic demand will increase them. The behavior Of eXport price depends upon whether the marginal cost curve is horizontal or upward leping and upon the elasticity of foreign demand for these exports. With respect to price, it is interesting to note that most discussions--as well as most empirical work--treat price as an exogenous variable. Dunlevy (1980) is the first to insist that even in the context Of this debate both price and quantity should prOperly be treated as endogenous variables. .muwoaxm wow panama oUmmHm Emuwfimfi ESE. mmmfl “Hope: 933% OLE .nIN 99me l9 xv H08 NOS _ _ to News as He . . .. u u . _ .. _ . . _ . HU . . . . . _ . _ _ NU u. _ . . . ._ _ . u H _. _ . . . _ ._ . . Ox . _ . .m 11111 III “I.II . . . _ . II IIIIII m _ . m. _ n . _ __ . . . _ I- I- lllll . . . . b H... IdI -.IIII n . . HOS .. . n 1111 I--- I 1| 11.. III..- . as -... ..-- ".--- --I a uuuuu 1 I IIIIIIII IIII IIII ....I. ma me me m wouoom flog 3v _ w 83mm 03888 A3 m. 20 This sub-section (2) Model of Non-Clearing Markets. consnnns models of non-clearing markets, a category which includes the eXplanations of eXport behavior which assume the existence of excess demand and the non-availability of twice as a rationing mechanism. Consider the domestic market and the eXport market for any commodity. The two interesting cases are, first, excess demand in both markets and, second, excess demand in the eXport market but not in the home market. If there already is excess demand in both markets, a rationing rule was clearly required in order to determine the degree to which the two sets of demands were tO be satisfied. If domestic demand should now change, output can be (re-)allocated according to the existing rationing rule. If there is excess demand in the eXport market only, it is necessary to consider separately the case of a decrease in domestic demand and the case Of an increase If domestic demand decreases, the in domestic demand. quantity of exports will increase. lfi‘domestic demand increases, however, there will be excess demand in both markets; and once again a rationing rule is required in order to determine whose demands shall be satisfied. Three rationing rules have been proposed thus far and none is entirely satisfactory. All three appeared in Ball LP961). One suggestion is that the percentage Of a firm's >roduction that is eXported is not permitted to fall below 21 somenunimum. This rule cannot tell how exports reSpond to acmmnge in domestic demand as long as the eXport percentage is above the required minimum. If the eXport percentage is equal to the desired minimum, an increase in domestic demand will have to be ignored (in order to prevent a constant amount Of eXports from becoming a smaller portion of a greater total output) or will be the cause of a corresponding increase in exports as the firm tries to maintain the appropriate ratio of export sales to domestic sales. This rule predicts the opposite of what the DPH predicts. A second suggestion is that a minimum absolute level Of eXports is maintained. By itself this rule does not bring any determinacy into the problem, but it could be used (serving, for example, as a constraint) in conjunction with another rationing rule. Thus, some other rule could be {killowed for the rationing of output provided that exports were not allowed to fall below some arbitrary minimum. The third possible rationing rule is that the domestic market is .satisfied first. It implies that there cannot be excess. Idemand imithe domestic market as long as the export quantity is greater than zero. It is this third rule that has received most attention, and therefore, it will be considered flnther. ldhy might suppliers give preference to domestic caustxnners when there is excess demand in the eXport market? 22 One possibility is that a higher unit profit on domestic sales induces producers to take care of the home market first.2 Another possibility is the force of habit. The existence Of search costs can make habits, traditions, and customs economically more efficient than the alternative of re-evaluating all business relationships each period. If customers grant a producer the privilege of being a regular supplier, the producer may find it in his long-run interest to satisfy the short-run changes in demand on the part of these traditional customers, even at the eXpense of temporary reductions in other sales. It is necessary, of course, to show that these well-established business relationships are more likely to be part of the domestic market than the eXport market. Although one can appeal to the costs of transportation, the difficulties of communication, and the general barriers created by cultural differences, it is still possible that for a particular firm, the force-of-habit eXplanation would imply that the export market rather than the domestic market is favored. A final possibility is patriotism or nationalism. Just as buyers are Often encouraged to buy from local suppliers, 2This eXplanation is implicit in Ball's (1961) contention that the price is likely to exceed the eXport price. Others who have picked up on the theme of profitability include Ball, Eaton, & Steuer (1966), COOper, Hartley, & Harvey (1970), Henry (1970), and Winters (1974). 23 prmhkmrs can feel obligated during periods of full capacity OpermnOns to take care of local requirements first.3 (3) "True"-Price Models. A third group of theories of eXpmfi;behavior is what may be called "true"-price models. These models claim that in addition to the price of a cmmmxfity (even when adjusted for inflation) buyers consider mnflifactors as credit terms and delivery lags in making purchase decisions. Certainly this is important in the case Of large ticket items such as aircraft or machinery, as evidenced by the current controversy over eXport credit Both of these factors can be viewed as part of subsidies. a seller the complete price. As domestic demand declines, may prefer to Offer more liberal trade credit rather than As domestic demand increases, change the quoted price. credit terms may become more stringent. Alternatively (or sinnfiltaneously) the market adjustment might occur by :fluctuations in the waiting time that the buyer has to ‘tolerate.“' During periods Of high demand, the customer 3'These last two eXplanations are the ones apparently favored by Henry (1970). lprhis issue was addressed specifically by Steuer, Ball, 8: Eaton (1966) in their investigation of the effect of The waiting times on eXport orders for machine tools. concept has also been used by Brechling & Wolfe (1965), Artus (1970), and Winters (197A). The most Smyth (1968L thorough treatment to date can be found in Greene (1975). 2” either receives the merchandise later than desired or, in order to assure timely delivery, commits himself to a purchase decision earlier than usual. This suggests that the true eXport price is indeed fluctuating in response to changes in domestic demand, and that these fluctuations allow markets to clear. But the fluctuations are occurring in the credit term or time lag component rather than in the direct monetary component of the price. (A) "True"-Quantity Models. The final group of microeconomic explanations Of eXport behavior is what may be called the "true"-quantity models. This category includes explanations that rely on a broadened concept of what is actually being purchased when one unit of a commodity changes hands. A purchase of an automobile usually includes the acquisition of a property right in the form of a warranty that covers major repairs for a specified period of time. Convenience and location of a dealer's repair shOp, perhaps staffed with competent personnel, is also a part of the purchase price. A customer buys more than just a commodity; he also "buys" a service and repair facility, the availability of spare parts, the use of temporary replacements, an exchange and refund policy, and professional consultations with the supplier's technical staff. As domestic demand increases, the producer may choose to reduce some of these ancillary 25 smwdces Thus the "quantity" of the good being eXported mayrun.be recorded as having changed, but indeed the fineigngmrchaser is getting "less" of the gOOd. The imporhnme Of these auxiliary facilities was recognized by Artus<fl970) and Henry (1970) but not incorporated formally The theoretical treatment of Ball (1961) into a model. Although it incorporates what he calls "selling services." canlneargued that the true-quantity models are simply a variation Of the true-price models, the approach taken here is to distinguish between the two. This distinction is helpful because in the true-price models, the commodity being purchased remains physically the same while the date of delivery changes or the date Of payment (along with interest eXpense) changes. In the true-quantity models, different "versions" of the commodity are being purchased-- a bicycle with a repair facility one mile away for example, rather than a bicycle with a repair facility 250 miles away. The Current Status _o_§ the Debate. T7”: best discussions of the Demand Pressure Hypothesis ~ I. re contained in the writings of COOper, Hartley, 8: Harvey COOper, Hartley, & Harvey 1970) and Dunlevy (1980). ~ovide a comprehensive analysis Of a neoclassical profit- Iximizing producer of a homogeneous product being sold in 0 distinct markets. They examine the various combinations price-maker's and price-taker's markets in cases of 26 increasing, constant, and decreasing marginal costs. What appears in Chapter 3 below draws heavily on the microeconomic, marginalist foundations that they have deveIOped. In addition, they investigate the implications of three other models Of the firm full-cost pricing, sales maximization, and satisfying behavior. \ Dunlevy's contribution is to point out that much of the discussion has focused on unnecessary issues; they arose because of failure to recognize, or an inability to cope with, the endogenous character Of export prices. From the beginning the posited inverse relationship between exports and domestic demand has been justified--sometimes eXplicitly, sometimes implicitly--on the basis of non- clearing markets. It may indeed be the case that markets do not clear (rapidly), and it may be true that the existence Of non-clearing markets combined with certain types of rationing rules will produce an inverse relationship between domestic demand pressure and exports. But this reliance on non-clearing markets is unnecessary. Rather, the Demand Pressure Hypothesis is a logical consequence of market- clearing behavior, where, for example, a rightward shift in the domestic demand function will cause a leftward shift in the eXport supply function--the usual result (depending upon elasticities) is that the export price rises and the quantity Of eXports decreases. In such a framework, both 27 pricue and quantity are endogenous variables. As Dunlevy poiths out, it is not an interesting question to ask how nuxfld of a change in quantity is induced by the price change Retina“, the issue now is to identify the cause of both the pricxa change and the quantity change; that is, the theoretical task is to identify the determinants of supply enui of domestic demand (these two sets together become the determinants of eXport supply) and the determinants of demand for eXports. Changes in any of the non-price determinants will have effects on eXport price and export quantity. The empirical task relevant to the Demand Pressure Hypothesis is to measure these effects as caused by changes in the determinants of domestic demand. Dunlevy presents a structural model where the quantity Of a country's total eXports demanded by the Rest of the World is a function Of the home country's eXport unit value index, a unit value Of all world eXports (including eXports Of the home country), and the aggregate value of world eXports (less the home country's imports). In turn, the quantity Of exports supplied is a function of eXport unit value, domestic price, domestic economic capacity, and umasures of domestic capacity pressure. The equilibrium rmnflrements is that the quantity Of eXports supplied be equal to the quantity demanded. 28 Having established the theoretical attractiveness Of the simultaneous equation approach, it is worthwhile carrying the analysis one step further than does Dunlevy. And this is to point Out that two Demand Pressure Hypotheses can be deduced from this simultaneous system. The first is the more general one; it would consider any change in the domestic demand function as the initiating shock which affects exports of a particular commodity. This version claims that it is the domestic demand pressure prevailing in one particular industry that affects the eXports of that industry.5 The second version is a special case of the first: rather than considering all the determinants of domestic demand, it considers only the income determinant. As industrial production, Gross National Product, and Personal Income change, the domestic demand function for an individual commodity will shift. This, in turn, will lead to a shift in the eXport supply function.6 Although Dunlevy is apparently the first to use a simultaneous equations approach to the Demand Pressure Hypothesis, others--for example, Morgan & Corlett (1951), 5As will be seen below, this is consistent with the use by Artus of industry-specific capacity utilization rates. 6This version better captures the flavor Of most of the discussion in the literature. It is consistent, for example, with the use of the economy-wide capacity utilization rate as an indicator of domestic demand pressure. 29 Bergstrom (1955), Swamy (1966), and Goldstein & Khan (1978)--have used a similar technique in their models Of eXport supply and demand. III. Empirical Development 5221: Demand Pressure Hypothesis. The previous section surveyed the contributions to the [Mnnand Pressure Hypothesis from the point of view of ecxmnomic theory. This section considers the empirical «contributions. With the notable exception Of Ball (1961), who is interested solely in deriving theoretical implications rather than in any empirical research, the papers to be reviewed are to a large extent the same as those in the previous sections. The reason for this is that the theoretical develOpment Of the Hypothesis has taken place in the empirical literature. First to be examined are three studies concerned mainly with issues other than the DPH, but which included in their eXport function a variable that represented the pressure of domestic demand on capacity. Next to be considered are those tests of the DPH that used single equation models, to be followed finally by the available studies that employ a simultaneous equations model to test the Demand Pressure Hypothesis. A. Three Estimates of Exports Functions. To be considered here are the studies by Renton (1967), Donges (1972), and Batchelor & Bowe (1974). In an attempt to forecast United Kingdom eXports Of mmnflactures to industrial countries, Renton (1967) 30 31 esthnates eight differently Specified equations, each having ‘the \Ialue Of U.K. exports of manufactures as the dependent VEHfiiable. The independent variables in three of them ituzlude a variable meant to measure the pressure of domestic denmnui, proxied by the ratio Of an index of seasonally adjusted U.K. manufacturing production to the trend value of this index. Fitted to quarterly data for 1956.1 through 1966.3, all three equations show a significant negative coefficient Of the pressure variable. Donges (1972) analyzes the demand and supply factors that affect Spain's eXports of manufactured items. He fits a single equation model to annual data for the period 1951- 1969 for Spanish eXports (dollar value) of total manufactures and for each Of twenty manufacturing industries. The variable that measures domestic demand pressure is a specially constructed series Of capacity utilization rates for each industry (used in the single- industry equations) and for all industries (used in the total manufactures equations). In the series, capacity output is measured according to the Wharton method (Klein & Summers, 1966). For three industries (processing food, leather and leather manufactures, and non-metallic mineral mmuMacturers), the coefficient of the capacity utilization rateis negative'and significant at the 10% level or better; fiN'twO industries (tobacco and chemicalS), it is 32 sigxrificantly positive at the 10% level. Batchelor & Bowe(fl974) develop a general equilibrium nuddel. for forecasting U.K. international trade as an aid to investment planning in the waterborne shipping industry. 'They lee two-stage least squares (ZSLS) to estimate U.K. eXpomW; demand and U.K. eXport supply equations for forty- five commodities. In the export supply equation, price is the left-hand variable, eXport so that a positive coefficient on the pressure variable (which is the ratio of U.K. output for the industry to its trend value) would support the DPH. For the following six industries, there is such a positive coefficient, significant at the 10% level or better: (1) sugar and sugar preparations, (2) beverages, not elsewhere specified, (3) organic chemicals, (A) road vehicles, n.e.s., (5) motor cars, and (6) glass and pottery. B. Tests of the Demand Pressure Hypothesis. The standard test of the DPH reported in the literature is a single-equation model that regresses the value or the volumecfi‘eXports against several eXplanatory variables chosmnfrom the following list: eXport price, world prices, Ikmmstic prices, world economic activity, domestic economic mfldvity, and a measure of domestic demand pressure. The following discussion considers, in turn, three fundamental aspects of these studies.7 First, what is being explained; thatis, what is the dependent or left-hand variable in the 33 equation? Second, what variable is chosen to represent cknnestic demand pressure and how does it enter the model? Thirwi, how does the author interpret the econometric results? Do these results support or fail to support the Demand Pressure Hypothesis? (1) The Dependent g: Left-Hand Variable. In all the studies, the dependent variable is some measure of eXport performance, with the different authors selecting differing degrees of aggregation, making different choices with respect to the volume-or-value question, and employing different indicators Of "performance." Consider first the degree of aggregation. In a graphical (non-econometric) study, Brechling & Wolfe try to eXplain the U.K. trade gap (imports minus eXports) in current prices. In the econometric studies, Winters and Dunlevy use total exports; Ball, Eaton, & Steuer and Smyth use total manufacturing exports; and COOper, Hartley, & Harvey, Artus, and Henry use individual commodities as the left-hand variable. All studies use the UJC as the home country. In addition, Dunlevy also tries tO eXplain total eXports from the UAL; and Henry runs individual regressions for thirteen commodities exported 7The discussion covers the following studies: Brmflfling & Wolfe (1965), Ball, Eaton, & Steuer (1966), Smymi(1968), COOper, Hartley, & Harvey (1970), Artus (19HD, Henry (1970), Winters (197A), and Dunlevy (1980). 3“ from Belgium and for eight commodities from the U.S. as well as for five commodities from the U.K. .Although much of microeconomic analysis speaks in terms of anantity, no study thus far has used an actual quantity, such as tons, gallons, carloads, or dozens, to measure the dependent variable, perhaps because the data sets have been too highly aggregated. Instead, the volume of eXports is o, dATC/dQ > 0. Equation (3-1A) permits the coefficients on all the terms, not just the domestic demand terms, to change when capacity output is reached. Two export supply functions which assert that all the coefficients except the domestic demand coefficient are independent Of capacity utilization rates would require the use of interactive terms and can be derived in the following manner. Let DUMCAP be a dummy variable that equals one when ATC and MC are both upward leping and equals zero 62 otherwise. Define the variables PHCAP and YHCAP as follows: PHCAP YHCAP PH‘DUMCAP, YH'DUMCAP. The variable YHCAP takes on the value of YH whenever the Average Total Cost curve is upward SIOping (that is, when the firm reaches capacity) and takes on the value zero otherwise (that is, below capacity). The same applies to PHCAP. A modification of Equation (3-14) would then be the following: QXSzQXS(PX,PHCAP,YHCAP,PLAND,PL,PK). (3-15) Equation (3-14) is more general than Equation (3-15): Equation (3-15) states that the relationship between QXS and PK (that is, dQXS/dPK) is the same whether the firm is at capacity or below, while Equation (3-1”) would permit the relationship to change once capacity output is reached. A final eXport supply function borrows from three develOped earlier: Equations (3-10), (3-1”), and (3-15). Domestic demand is part Of Equation (3-10) for all Observations; and it is part Of Equations (3-1“) and (3-15) only if marginal cost is rising (that is, only if domestic demand is high). Domestic demand, however, enters the next export supply equation in such a way that different responses to home income can be hypothesized when marginal cost is constant than when marginal cost is rising. That is, different reSponses to home income can be hypothesized when production is beneath capacity than when production is 63 at or beyond capacity. In this version, not only are PH and YH in the eXport supply function but so are PHCAP and YHCAP. Recall that YHCAP is equal to zero when the firm is producing beneath capacity and becomes equal to YH at capacity output. This technique allows us to test for an "average" relationship between the quantity of exports supplied and home demand (this will be the coefficient on YH) and also to test for the change in this relationship when the firm is producing at high capacity (this will be the coefficient on YHCAP). This export supply function can be written as follows: QXS=QXS(PX,PH,PHCAP,YH,YHCAP,PLAND,PL,PK). (3-16) III. The Complete Models A complete model must consist of demand and supply functions and an equilibrium condition. The demand-for- eXport function, derived earlier as Equation (3-6), can be forwarded without change as Equation (3-12): QXD = QXD(PX,XRINDX,YROW). (3-12) In turn, there are five eXport supply functions--Equations (3-10),(3-11),(3-14),(3-15),annl(3-16). Equation(3-11) claims that there is no relation between eXports and domestic demand; while the other four assert that such a relation does exist, with each specifying a somewhat different effect on eXports from changes in domestic demand. The equilibrium condition is that the quantity of eXports supplied be equal to the quantity demanded: 6A QXS : QXD. (3-13) Combining the demand for export function with each Of these export supply functions, under the equilibrium condition, yields five simultaneous-equation models of eXport behavior. The models--denoted A, B, C, D, and E--are presented in Figure 3-6. In all five models the equilibrium quantity of eXports (OX) and the equilibrium eXport price (PX) are dependent variables; the other variables are independent. The reduced form equations for these models are presented in Figure 3-7. For convenience the variables are listed alphabetically and defined in Figure 3-8. 65 Model A, Constant Marginal Cost: Supply: QXS=QXS(PX,PLAND,PL,PK) (3-11) Demand: QXD=QXD(PX,XRINDX,YROW) (3-12) Equilibrium: QXS=QXD (3-13) Model B, Rising Marginal Cost: Supply: OXS=QXS(PX,PH,YH,PLAND,PL,PK) (3-10) Demand: QXD:QXD(PX,XRINDX,YROW) (3-12) Equilibrium: QXS=QXD (3-13) Model C, Constant and Then Rising Marginal Cost: Supply: { QXS(PX,PLAND,PL,PK) QXS : { if dMC/dQ : O (3-1A) { QXS(PX,PH,YH,PLAND,PL,PK) { if dMC/dQ > 0, dATC/dQ > 0. Demand: QXD:QXD(PX,XRINDX,YROW) (3-12) Equilibrium: QXS=QXD (3-13) Model D, Constant and Then Rising Marginal Cost: Supply: QXS=QXS(PX,PHCAP,YHCAP,PLAND,PL,PK) (3-15) Demand: QXD=QXD(PX,XRINDX,YROW) (3-12) Equilibrium: QXS=QXD (3-13) Model E, Eventually Rising Marginal Cost: Supply: QXS=QXS(PX,PH,PHCAP,YH,YHCAP,PLAND,PL,PK) (3-16) Demand: QXD=QXD(PX,XRINDX,YROW) (3-12) Equilibrium: QXS=QXD (3-13) Figure 3-6. Simultaneous Equation Models: Models A, B, C, D, and E 66 Model A, Constant Marginal Cost: QX : QX(YROW,XRINDX,PLAND,PL,PK) (3-17) PX = PX(YROW,XRINDX,PLAND,PL,PK) (3-18) Model B, Rising Marginal Cost: QX = QX(YROW,XRINDX,PH,YH,PLAND,PL,PK) (3-19) PX = PX(YROW,XRINDX,PH,YH,PLAND,PL,PK) (3-20) Model C, Constant and Then Rising Marginal Cost: 1. When marginal cost is constant: OX 2 QX(YROW,XRINDX,PLAND,PL,PK) (3-17) PX = PX(YROW,XRINDX,PLAND,PL,PK) (3-18) 2. When marginal cost is rising: QX : QX(YROW,XRINDX,PH,YH,PLAND,PL,PK) (3-19) PX = PX(YROW,XRINDX,PH,YH,PLAND,PL,PK) (3-20) Model D, Constant and Then Rising Marginal Cost: QX(YROW,XRINDX,PHCAP,YHCAP,PLAND,PL,PK) (3-21) PX(YROW,XRINDX,PHCAP,YHCAP,PLAND,PL,PK) (3-22) QX PX Model E, Eventually Rising Marginal Cost 0k>"a priori" statement on Marginal Cost initially): QX PX QX(YROW,XRINDX,PHCAP,YH,YHCAP,PLAND,PL,PK) (3-23) PX(YROW,XRINDX,PHCAP,YH,YHCAP,PLAND,PL,PK) (3-29) Figure 3-7. Reduced Forms: Quantity and Price Equations 67 PH The real home price of the eXportable commodity. PHCAP Interaction term. Equals the home price Of the exportable commodity when the capacity utilization rate is high and equals zero otherwise. PK The real price of capital. PL The real price of labor. PLAND The real price of land. PX The real price of the exported commodity. OX The quantity of exports (individual commodity). XRINDX The foreign exchange value of the domestic currency. YH Real home income. YHCAP Interaction term. Equals real home income when the capacity utilization rate is high and equals zero otherwise. YROW Real income in the rest of the world. Figure 3-8. Definitions of Variables. CHAPTER u DATA I. Introduction. This chapter describes the data used to estimate the models develOped in Chapter 3. Particular attention is paid to the handling of missing data and to describing some necessary adjustments to the statistics (detailed descriptions are presented in Appendix 1). In addition, some reference is made to data that were used as eXplanatory variables in preliminary experimentations but that were not included in the final formulations (a complete discussion is presented in Appendix 2). II. Trade Data. U.S.eXport data come from‘thelI£L Department of Commerce monthly fOreign trade report FT A10 U.S. Exports: Schedule E Commodity Groupingsl Schedule E Commodity py Country. These are 7-digit commodities classified according to Schedule B prior to 1978 and according tO Schedule E beginning in 1978. Commodities were selected at random, subject to the constraints of data availability and the desire to have representation from each Of SITC Sections 5, 6, 7, and 8. Of the forty-seven commodities for which data had been originally collected, sixteen were eliminated from the study because Of the difficulties Of matching the pre- 68 69 1978 Schedule B numbers with the new Schedule E numbers. For those commodities that could be carried forward to the end of 1979, concordance was Obtained by going from Old Schedule B (which was based on SITC) to the new Schedule 8 (based on the Tariff Schedule of the U.SJ and then from new B to new Schedule E (which is SITC-based).l3 The quarterly data Span the period 1965.1 to 1979.A, for a total of sixty Observations. The commodities are identified here by their Schedule E numbers. Figure A-1 lists the commodities along with their schedule E numbers. The right hand column shows the previous number, with which the new number was concorded on the basis of the descriptions and with the aid Of the concordance tables.l“ 13The desired eXport data are published in FT 410, a foreign trade report fO the Department of Commerce. Prior to 1978 commodities in FT 410 are classified according to Schedule B, which was SITC-based. Beginning in January 1978, the Schedule B numbering system changed, and the new Schedule B is now based on the Tariff Schedule of the U.S. A new schedule was created, called Schedule E, that is SITC-based. Schedule E is similar to, but not identical with, the Old Schedule B and now forms the basis for classification of commodities in FT 410. There is no concerdance table available that goes directly from old Schedule B to new Schedule E--that is, from Old FT 1110 to new FT 410. 1“Commodity NO. 525 6030 is missing an observation for 1965.3; this precludes finding quarterly data for 1965.“. Rather than discarding the observations for 1965.1 and 1965.2 and beginning the regressions with 1966.1, Observations for 1965.3 and 1965.4 were approximated by linear extrapolation. (3‘ 0‘ 0‘ kn U1 *1 NJ 1: O» U1 U) ... ...: (llLA) Ch 0‘ (1" (Y) I\) R) 0‘ 0“ I (7‘ (J) ‘— Schedule E Number Brief Description (Schedule B) Previous Numbers 525 553 588 641 671 673 682 682 684 684 686 691 691 695 695 696 716 721 723 724 742 Figure 4-1. 6030 2000 3060 1000 2000 2005 2160 2400 2140 2420 3220 1020 2020 3140 11145 0340 4042 2220 4052 3920 4026 Iron oxides and hydroxides, pigment grade Printing inks Rubber cement Newsprint Pig iron, including cast iron Concrete reinforcing bars (also includes since 1978: Copper alloy wire, bare Copper and copper alloy powders and flakes Aluminum and aluminum alloy wire, except insulated Aluminum and aluminum powders and flakes (also includes since 684 2440.) Zinc and zinc alloy sheets, plates, and strip Door and door and window sash, frames, and molding and trim, of iron and steel Door and door and window sash, frames, and trim, of aluminum Hacksaw blades, hand and power (beginning in 1979.1 this number Splits into 693 3139 and 693 3193: both are included here.) Twist drills and drill bits, metal cutting. Safety-razor blades Motors, AC, polyphase—induction, not over 20 HP Combines, self-propelled Dozers, for mounting on tractors Needles, sewing machine Centrifugal pumps for liquids, single- stage, single-suction, close-coupled, under 2 inch outlet 1978.2: Description Of Commodities Tested 673 2010.) 0210 2420 2150 0130 2400 2010 2010 2140 2440 0320 1034 2010 4244 3070 2120 2121 513 5030 (1965-69) 513 5320 (1970—77) (1965-77) (1965-77) (1965-77) (1965-77) (1965-77) (1965-77) (1965-77) (1965-77) (1965-77) (1965-77) (1965-77) (1965-77) (1965-77) (1965-77) (1965-73) (1974-77) Figur 71 Number Brief Description Previous Numbers 743 1035 Air compressors, stationary, over 100 HP 719 2220 (1965-77) 751 1040 Typewriters, standard, non-portable, electric, new 714 1010 (1965-77) 762 0040 Radios, household-type, without phonograph 724 2010 (1965-77) 775 4030 Shavers, electric 745 0410 (1965-77) 775 7520 Vacuum cleaners, electrO-mechanical 725 0320 (1965-77) 775 8625 Toasters, automatic, electric, household type 725 0520 (1965-77) 884 2220 Sun or glare glasses and sun 861 2010 (1965-69) goggles 861 2210 (1970-77) 885 2020 Clocks, electric 864 0120 (1965-69) 864 0320 (1970-77) 891 0945 Tape, pressure-sensitive plastic 893 0045 (1965-77) 895 2115 Ball-point pens and ball-point pencils 895 2120 (1965-77) Figure 4-1. Description of Commodities (cont.) wor pre A. III. Macroeconomic Data. Five basic categories of macroeconomic data are needed: world output, exchange rates, U.S. output, domestic demand pressure, and factor prices. A. World Output. Data furnished by the United Nations Statistical Office were used to construct a time series Of Value Added by the rest-of-the-world (that is, the world excluding the United States) in mining, manufacturing, electricity, gas, and water. The series is denoted VAROW and was computed in the following way. First, the World value-added figures are constructed by multiplying quarterly index numbers of world industrial production times the value-added by the world in 1975 (eXpressed in 1975 U.S. dollars). The result is a quarterly time series for world value added, eXpressed in 1975 dollars. Second, use of the same procedure for the United States yields quarterly data for U.S. value added, also eXpressed in 1975 Ufih dollars. Finally, by subtracting U.S.‘value-added from world value-added, the time series VAROW is constructed. VAROW enters the regressions denominated in millions of 1975 U.S. dollar.15 15The countries that comprise the "world" do not remain constant throughout the sample period. In 1965, for example, the world excludes China (Mainland), North Korea, North Viet-Nam, U.SJLR., and Eastern EurOpe. In 1968 and 1969, the world excludes China (Mainland), 72 De as 99 De! dUI 73 B. Dollar Exchange Rate. To measure the effective exchange rate of the U.S. dollar, a time series was constructed (and denoted as the variable XRIMF) from the International Monetary Fund's series on effective exchange rates. XRIMF is defined to indicate a rise in the value Of the dollar as XRIMF increases. The series is derived from the Fund's Multilateral Exchange Rate Model (MERM) and is published as line "amx" in the Fund's International Financial Statistics. The weights used in constructing the series are generated by the Fund's Multilateral Exchange Rate Model and represent the model's estimate of the effect on the U.S. trade balance Of a one percent change in the dollar value of one of the other currencies.16 It is an arithmetic average for the period rather than an end-Of-period value and is constructed as an index in which the par values in May 1970 are set equal to 100. Beginning with the third quarter of 1979, the base period was changed to the average market exchange rates during 1975. The 1975-based series had to be converted into Mongolia, Democratic PeOple's Republic of Korea, and the Democratic Republic of Viet-Nam. By 1977 and 1978, the world excludes Albania, China, Mongolia, Democratic People's Republic of Korea, and Socialist Republic of Viet-Nam. 16A description of the series can be found in IMF (1980). SET: maj: inde this era EEVa and 1 74 a 1970-based series for the last two Observations (1979.3) and 1979A) in the regression analysis. This was done in the following way. Back values of the 1975-based index are furnished by the IMF beginning with 1976.1. For the period 1976.1 through 1979.2, therefore, the values for both the 1970-based series and the 1975-based series are available from the IMF. During this period, the average relationship between the two series is: 1970-based index equals 0.8352398 times 1975-based index. That conversion factor was used to obtain 1970-based values for 1979.3 and 1979.4 from the 1975-based index. The effective exchange rate series gives the value of the dollar vis-a-vis twenty other major currencies. The published series begins with the value for 1972.1. In order to get a series going back to 1965, the Fund's index was used for all observations from 1972.1 forward and this index was approximated backwards into the Bretton Woods era by taking into a account the devaluations and revaluations of the currencies Of Canada, Germany, France, and the United Kingdom. The method used to accomplish this is described in Appendix 1. C. U.S. Output. U.S. Gross National Product at 1972 prices was used to measure U.S. output. This variable is given the name GNP72$. In some preliminary work, the Federal Reserve Syste: this I Reserx each i Reserv proces each 0 0Ver t rate It analys II ProduC. Statist reel tI Series labor c real te 75 System's Index of Industrial Production was employed, but this was abandoned in favor of GNP72$. D. Demand Pressure. Domestic demand pressure is measured by the Federal Reserve System's capacity utilization rates as published in each issue of the Federal Reserve Bulletin. The Federal Reserve measures capacity as a percent of 1967 output. Output is converted to an index (1967 = 100), and then the capacity utilization rate is output divided by capacity. Three of the Federal Reserve System's series were tried: total manufacturing, primary processing, and advanced processing. Because the three are so highly correlated with each other, there is no significant advantage in using one over the other two. Consequently the capacity utilization rate for total manufacturing was used for all the regression analysis. E. Input Prices. In order to measure the prices Of the factors of production, two nominal series from the Bureau of Labor Statistics, Employment and Earnings, were converted into real terms by dividing by the GNP deflator. To Obtain a series measuring the price of labor, the series called "unit labor cost, private sector, non-farm" was converted into real terms and called PL. TO obtain a series measuring the prices of both capital and land, the series called "unit non-l into and w descr see A, 76 non-labor payments, private sector, non-farm" was converted into real terms and denoted PK. Several other measures of factor costs are published by the Bureau of Labor Statistics and were tried in preliminary investigations. For a description of these and their relationship to PL and PK, see Appendix 2. CHAPTER 5 RESULTS OF REGRESSION ANALYSIS I. Estimation Procedure. Two modifications have to be made to the reduced form equations Of Figure 3-7. First, the price Of land and Of capital are combined into one term--unit non-labor cost (PK). This was dictated by data availability. Second, the domestic price of the home country's eXported commodity (PH) is eliminated from all equations. This was dictated by inadequate concordance between export classification (Schedule E) and domestic classification (SIC) Schemes, particularly at the 7-digit level. The reduced form equations that are to be estimated are presented in Figure 5-1. The generalized notation used in Chapter 3 has in several instances been replaced by notation that is more descriptive of the actual data series employed. Thus, for Rest-Of-World Income the series used is Value-Added in the Rest of the World, and therefore, YROW becomes VAROW. The exchange rate index used in the regressions comes from the IMF, and thus XRINDX is replaced by XRIMF. Gross National Product in 1972 dollars is used as the measure of home income, and therefore, GNP72$ replaces YH. For each commodity, four different capacity utilization rates (83, 85, 86, and 87%) are tested as working definitions Of full capacity, and thus the variable described earlier as YHCAP 77 be YH wh (G No 86‘. re: ut: the Re: KII COP. Utj W01- rat the Dec Six 78 becomes four different variables: YHCAP83, YHCAP85, YHCAP86, and YHCAP87. Model B degenerates into Model A whenever the coefficient on the home income variable (GNP72$) is not significantly different from zero; that is, Model A asserts that domestic demand pressure does not affect eXports. In order to estimate Models C and D, a satisfactory method has to be found for determining when the economy is producing under conditions of constant unit costs and when it is eXperiencing rising marginal and average costs. The concept used here is the capacity utilization rate. Although capacity is defined theoretically in the Klein sense as the output at which unit cost begins to rise, it is recognized that the published Federal Reserve capacity utilization rates do not follow this definition. Part Of the task of the regression analysis is to find the Federal Reserve capacity utilization rate that corresponds to the Klein view of a 100% capacity utilization rate. For each commodity, four different Federal Reserve capacity utilization rates (83, 85, 86, and 87%) are tested as working definitions of Klein's 100% capacity utilization rate. This technique furnishes a link between the theoretical treatment Of Chapter 3 above and the practical necessity Of employing the data that are available. Of the sixty observations, thirty-seven have a capacity utilization util high capaI Obse: these tests coeff Sourc is an or of relat ”0 Fe: in the Zero E that E equal test. Zero ( test 1 237 79 rate Of 83% or higher; thirty-one have a capacity utilization rate Of 85% or higher; twenty-four, 86% or higher; and sixteen, 87% or higher. The next higher capacity utilization rate--88%--would reduce the number of Observations in the high-pressure category to nine, and these would be the first nine quarters Of the sample. For interpreting the regression results one usually tests for the statistical significance Of the estimated coefficients. The usual procedure is as follows.17 The theory that one is trying to "prove" or "confirm" is the source of the alternative hypothesis. The null hypothesis is an implication of the current generally accepted theory or Of an eXplanation which claims that there is no relationship among the variables being studied. If there is no relationship among the variables, the true coefficients in the equation are equal to zero, and any deviations from zero arise from random chance. If the null hypothesis is that B equals zero and the alternative is that B is not equal to zero, then the apprOpriate test is a two-tailed t- test. If the alternative hypothesis is that B is less than zero (rather than simply not equal to zero) the apprOpriate test is a one-tailed t-test. 17See, for example, Kmenta (1971), p. 114 and pp. 236- 2370 apprOpr t”°~tai coeffic; 80 Writers on the Demand Pressure Hypothesis have taken the view that the DPH is the alternative hypothesis and that some null hypothesis shall be accepted (more accurately "not rejected") unless the evidence is overwhelmingly in favor of the DPH (in which case the null hypothesis is rejected in favor Of the DPH). The same procedure will be followed in this study. Some discussion is necessary, however, concerning the apprOpriate t-test-- whether it Should be one-tailed or two-tailed. The DPH claims that increases in domestic demand will cause the quantity Of eXports tO decrease and the price of eXports to increase. Thus, for the quantity equation the eXpected coefficient on the pressure variable is negative and for the price equation this eXpected coefficient is positive. It seems apprOpriate to use a one-tailed t-test for this particular class of coefficients. There is no intention here of testing any particular hypotheses regarding the other explanatory variables; that is, the DPH itself, which is the focus of this study, has nothing to say about the relationship between the eXport quantity or price and the non-pressure independent variables. In light of this the apprOpriate test, if any, for these variables would be a two-tailed t-test with a null hypothesis that the coefficient is zero and an alternative hypothesis that the coefficient is not zero. ”Y! I D >.; I": H H 81 Model A, Constant Marginal Cost (the Null Hypothesis): QX PX QX(VAROW,XRIMF,PL,PK) PX(VAROW,XRIMF,PL,PK) Model B, Rising Marginal Cost: QX PX QX(VAROW,XRIMF,GNP72$,PL,PK) PX(VAROW,XRIMF,GNP72$,PL,PK) Model C, Constant and Then Rising Marginal Cost: 1. When marginal cost is constant (low capacity utilization): QX PX OX(VAROW,XRIMF,PL,PK) PX(VAROW,XRIMF,PL,PK) 2. When marginal cost is rising (high capacity utilization): QX PX QX(VAROW,XRIMF,GNP72$,PL,PK) PX(VAROW,XRIMF,GNP72$,PL,PK) Model D, Constant and Then Rising Marginal Cost: QX PX QX(VAROW,XRIMF,YHCAP,PL,PK) PX(VAROW,XRIMF,YHCAP,PL,PK) Model E, eventually Rising Marginal Cost (NO "a priori" statement on Marginal Cost initially): QX 'PX OX(VAROW,XRIMF,GNP72$,YHCAP,PL,PK) PX(VAROW,XRIMF,GNP72$,YHCAP,PL,PK) Figure 5-1. Reduced Forms TO Be Estimated. variabl the coe tailed respect; used to 82 II. Discpssion of Resglts. ,_ _ -..—.— _-_ _ _ __ __ A- 9158.11181- In the tables containing the regression results (Tables A-1 through A-80), the t-ratios for the pressure variables are marked with one, two, or three asterisks if the coefficient is statistically Significant in a one- tailed t-test at the 10%, 5%, or 1% probability level, respectively. In an analogous fashion, the "#" symbol is used to denote statistical significance in a two-tailed t- test. Table 5-1 summarizes the results. The column entitled "Number Of Tests" requires some eXplanation. An estimated regression equation is considered a "test" in this Table when the equation comes form Models B, D, or E. A test from Model C requires two equations: one at low demand pressure and one at high demand pressure. The most common number of tests is eight: one from each of Models B, C, D, and E for the quantity reduced form and one from each for the price equation. For several Of the commodities, interesting (sometimes conflicting) results showed up at more than one definition Of full capacity (that is, at more than one cut-Off value for the capacity utilization rate). These are the commodities for which more than eight tests are reported. Altogether, there are 89 cases in which the null hypothesis can be rejected in favor Of the Demand Pressure 83 Hypothesis at the 10% level or better and 216 cases in which the null hypothesis cannot be rejected. Of the 89 cases supporting the DPH, 47 come from the quantity equations and 42 from the price equations. In 19 cases (involving 12 different commodities), there is support for the DPH in both the quantity equation and the price equation of the same Model. Of the 31 commodities that are used for these 305 cases, there are only four commodities for which the null hypothesis cannot be rejected in any of the tests; that is, there is at least some support for the DPH in 27 out of the 31 commodities. The four commodities that lend no support to the DPH come from three different SITC Sections: 533 2000, printing inks, from Section 5; 684 2140, aluminum and aluminum alloy wire, not insulated, and 695 3140, hacksaw blades, hand and power, from Section 6; and 775 8625, toasters, automatic, electric, from Section 7. In the case of commodities from SITC Sections 5, 6, and 7, the effect of domestic demand pressure shows up more Often on the quantity Of exports than on the price of eXports; whereas in the case of Section 8, the effect shows up on quantity five times and on price eleven times.18 For the three commodities classified as 18Section 5 is chemicals and related products, not specifically provided for. Section 6 is manufactured goods classified chiefly by material. Section 7 is machinery and transport equipment. Section 8 is miscellaneous manufactured articles. 84 chemicals (SITC 5), four quantity equations and three price equations support the DPH. For the thirteen commodities in SITC 6 (Manufactured goods), support comes from nineteen quantity equations and thirteen price equations. And for the eleven commodities in Section 7 (Machinery), nineteen quantity equations and fifteen price equations support the Demand Pressure Hypothesis. The results of the regression analysis are discussed below according to SITC Section. The first three commodities are discussed in greater detail than the rest in order to familiarize the reader with the procedures employed and the format Of the tables in Appendix 3. 85 Table 5-1 Summary of Regression Results (1) (2) (3) (4) (5) (6) (7) Cannot Reject Support the Null Commodity Number lps 23H Hypothesis of Number Brief Description Tests 9X EX QX BX 525 6030 Iron oxides 14 3 3 4 4 533 2000 Printing inks 8 0 0 4 4 588 3060 Rubber cement _8 _l _9 _; .5 .32 1 .3. 1_1 12. 641 1000 NewSprint 12 4 2 2 4 671 2000 Pig iron 10 2 O 3 5 673 2005 Concr.reinforc.bars 8 2 1 2 3 682 2160 COpper wire 14 2 0 5 7 682 2400 COpper powder 8 O 1 4 3 684 2140 Aluminum wire 8 O O 4 4 684 2420 Aluminum powder 8 0 1 4 3 686 3220 Zinc sheets 8 3 1 1 3 691 1020 Steel door frames 8 3 3 1 1 691 2020 Aluminum door frames 8 2 0 2 4 695 3140 Hacksaw blades 8 O O 4 4 695 4145 Twist drills 8 1 2 3 2 696 0340 Razor blades .__8 _Q _2 _3 _§ m 12 11 1.9. 15 716 4042 AC motors 14 2 3 5 4 721 2220 Combines 9 O 4 4 1 723 4052 Dozers 8 2 1 2 3 724 3920 Sewing mach. needles 8 4 1 0 3 742 4026 Centrifugal pumps 12 2 1 6 3 743 1035 Air compressors 12 3 3 3 3 751 1040 Electric typewriters 8 1 O 3 4 762 0040 Radios 8 1 O 3 4 775 4030 Electric shavers 8 2 2 2 2 775 7520 Vacuum cleaners 14 2 0 5 7 775 8625 Toasters __§ _9 _Q _3 _3 109 19 15 31 38 (Continued on next page.) Nun he 885 891 895 equat thrOué Quantl MEHY p many number cannOt fail t. entries Column 86 Table 5-1 Continued (1) (2) (3) (4) (5) (6) (7) Cannot Reject Support the Null Commodity Number the DPH Hypothesis of Number Brief Descrlption Tests OX PX OX PX 884 2220 Sun glasses 16 0 3 8 5 885 2020 Electric clocks 12 0 2 6 4 891 0945 Plastic tape 8 1 2 3 2 895 2115 Ballpoint pens 13 _5 _fl _3 _3 2.0. .211 22.11 Total 305 4_ 32 107 109 Total 305 8 216 NOTE: CoTumn 3 indicates the number Of regression equations whose estimations are reported in Tables A-1 through A-80. Of these estimations, Column 4 shows how many quantity equations support the DPH, and Column 5 shows how many price equations support the DPH. Column 6 indicates how many quantity equations fail to support the DPH--that is, the number of quantity equations in which the Null Hypothesis cannot be rejected. Column 7 shows how many price equations fail to support the DPH. For each commodity, the sum of the entries in Columns 4, 5, 6, and 7 is equal to the entry in Column 3. PX: A II negatix ”'52 1. Zero at 87 B. Commodities from SITC Section 2 (Chemicals). Consider the first commodity--525 6030, Iron oxides and hydroxides, pigment grade. Some of the estimated equations suggest that the Demand Pressure Hypothesis holds for this commodity while other equations contradict such a conclusion. Equations 1 and 2 from Table A-1 test Model B in Figure 5-1 and are reproduced below. They were estimated with an auto regressive technique; the numbers in parentheses under the parameter estimates are t-ratios, while the F-statistic numbers are the degrees of freedom in the numerator and in the denominator. QX : 34300 + .0522 VAROW - 148 XRIMF - 6.41 GNP72$ (2.25) (.0265) (-3.71) (-1.52)* - 34.6 PL - 52.6 PK + ERR, (-O.426) (-1.14) Rho : 0.586, F(5, 53) = 4.58, R-Squared: .30, OH: 2.19. (5.55) PX = - 1840 - .0966 VAROW + 3.45 XRIMF + 0.846 GNP72$ (-1.60) (-0.793) (1.18) (2.66)*** +6.85 PL + 0.141 PK + ERR, (1.24) (.0441) Rho: 0.840, F(5, 53): 1.82, R-Squared: .15, ON: 2.36. (11.9) In the quantity equation, the coefficient on GNP72$ is negative as predicted by the DPH, and the t-statistic of -1.52 indicates the the coefficient is significantly less than zero at the 10% level. The price equation also supports the DPH: the coefficient Of 0.846 on the GNP72$ term is sigr that whic Whenel the QC equati eQUati Coeffi althOu Sign. negati, ModelI utiliza QUantit: UtiliZat 88 significant at the 1% level (1-tailed t-test). It indicates that the eXport price rises as real domestic GNP increases, which is what the DPH predicts.. Tests of Model D are given as Equations 3 and 4 in Table A-1. In Model D the coefficient on GNP72$ is forced to be zero whenever the capacity utilization rate is beneath a certain cut-off rate. The rates tested for all commodities were 83, 85, 86, and 87 percent. With this model, the specific rate chosen for reporting purposes was the one with the most significant t-statistic (the unreported results are available from the author). For this commodity (iron oxides) the reported version Of Model D uses 83% as the cut-Off rate. The variable YHCAP83 is equal to GNP72$ whenever the capacity utilization rate is 83% or greater and is equal to zero whenever the rate is less than 83%. The DPH predicts that the coefficient on YHCAP83 will be negative in the quantity equation and positive in the price equation. The estimated equations, reproduced below, do not support the DPH. The coefficient on YHCAP83 is insignificant in both equations; although in the price equation it does have the predicted sign. The estimated Model D equations suggest that the negative coefficient on GNP72$ in the quantity equation of Model B is not arising from the pressure Of high capacity utilization; the negative relationship between GNP72$ and the quantity of eXports must be occurring at ls! rates of capacity utilization. Likewise, Model B showed a positive re ent bet at sig: 0): : PX : This 35 5, whic Variabl indicatE iRCreaSe capture approfiche actually significar that there 89 relationship between GNP72$ and the price of eXportS over the entire sample period, but Model D shows that the relationship between these two variables is not statistically significant at high capacity utilization rates. The positive (and significant) relationship evident in Model B must be arising during periods of low capacity utilization. QX : 18600 - 1.95 VAROW - 105 XRIMF + 15.8 PL (1.24) (-1.42) (-2.63) (0.189) - 50.7 PK + 0.324 YHCAP83 + ERR, (-1.01) (1.42) Rho: 0.741, F(5,53)= 2.72, R-Squared: .20, DW=2.33. (8.48) PX = - 78.9 + .0877 VAROW + .0814 XRIMF + 1.94 PL (-.0777) (0.791) (.0286) (0.349) - 0.u22 PK + .00421 YHCAP83 + ERR, (-0.12u) (0.279) Rho: 0.886, F(5,53)= 0.218, R-Squared: .02, 0W: 2.42. (14.7) This assertion is further supported by the results of Model E, which includes both GNP72$ and YHCAP83 as eXplanatory variables. The coefficients on GNP72$ are significant and indicate that quantity falls and price rises as real GNP increases; but the YHCAP83 term, which was intended to capture the extra impact Of GNP on eXports as the economy approaches capacity, shows that the quantity of eXports actually rises (for a 2-tailed test, the t value of 1.76 is significant at the 10% level) with high demand pressure and that there is no apparent effect on prices. Model E equat OX PX:. [H 9O equations are as follows: QX : 32800 + 0.974 VAROW - 140 XRIMF - 8.43 GNP72$ (2.03) (0.455) (-3.33) (-1.74)# -22.3 PL - 53.0 PK + 0.435 YHCAP83 + ERR, (-0.267) (-1.11) (1.76)# Rho: 0.670, F(6,52)= 3.30, R-Squared: .28, DW: 2.21. (6.93) -1960 - 0.119 VAROW + 3.47 XRIMF + 0.967 GNP72$ (-1.66) (-0.921) (1.18) (2.67)### PX + 6.68 PL + 0.454 PK - .0123 YHCAP83 + ERR, (1.20) (0.140) (-0.777) Rho: 0.860, F(6,52)= 1.48, R-Squared: .15, DW= 2.38. (13.0) Equations 7, 8, 9, and 10 in Table A-2 are tests of Model C, which hypothesizes that the coefficient on the pressure variable is zero at low capacity utilization rates and is negative for the quantity equation and positive for the price equation at high capacity utilization rates. Equations 7 and 8 use only those Observations for which the capacity utilization rate is less than 83, and equations 9 and 10 use the Observations for which the capacity utilization rate is 83 or greater. The quantity equation for low capacity utilization rates is shown below, followed by the one for high utilization rates. QX = 22130 + 1.80 VAROW - 151 XRIMF - 5.53 GNP72$ (-0.993) (0.509) (-4.48) (-0.933) + 267 PL + 74.6 PK + ERR, (2.24) (1.23) F(5,17) = 12.6, R-Squared : .79. II U1 I\) 01 0X Similarl first fc capacity PX ...: f\) C o 1 ( PX:-g 91 (ax = 52500 + 6.17 VAROW - 116 XRIMF - 17.7 GNP72$ (4.34) (1.38) (-2.46) (-2.53)*** -120 PL - 96.4 PK + ERR, (-1.63) (-1.59) F(5,31) = 11.7, R-Squared : .65. Sirnilarly the estimated price equations are shown below, firrst for low capacity utilization rates and then for high cap>acity utilization rates: PX. = 1220 - 0.364 VAROW + 5.86 XRIMF + 1.04 GNP72$ (1.16) (-2.18) (3.70) (3.73)### - 9.57 PL - 8.99 PK + ERR, (-1.71) (-3.16) F(5,17) = 12.8, R-Squared = .79. PX : - 4350 - 0.535 VAROW - 0.648 XRIMF + 1.54 GNP72$ (-4.91) (-1.64) (-0.187) (3.00)*** +20.6 PL + 8.00 PK + ERR, (3.82) (1.80) F(5,31) = 11.8, R-Squared = .65. For~ this commodity the coefficients on the GNP72$ term in Mcxiel C behave almost exactly the way the DPH predicts. For the: quantity equation, the GNP72$ coefficient is not Significantly different from zero at low utilization rates and is -17.7 (with a t-ratio Of -2.53) at high rates. For the price equation the values are 1.04 (t: 3.73) at low rates and 1.54 (t: 3.00) at high rates; both t-ratios are significant at the 1% level. These mixed results for iron oxide--with support for the DPH coming from Models B and C but not from Models D and E-- can be reconciled somewhat if alternative cutoff rates for A-3 comp; HCAI coeft quan: to he behav equat equat less eQuaI Drags GNP7; the 0 add 0 CaDac negat hishe the c (t: q. the 92 can be reconciled somewhat if alternative cutoff rates for capacity utilization rates are tested. This is discussed next. Results for a cut-Off rate of 85% are shown in Tables A-3 and A-4. Equations 11, 12, 13, and 14 in Table A-3 are comparable to Equations 3, 4, 5, and 6. The coefficients on YHCAP85 are insignificant for all four equations. The coefficient on GNP72$ continues to be negative in the quantity equation but no longer significant, and it continues to be positive and significant in the price equation. An interesting feature Of the Model C estimations is the behavior of the GNP72$ coefficient in the low-pressure equations as the definition Of low pressure is changed. The equations in Table A-2 use a capacity utilization rate of less than 83% as the definition Of low demand pressure; the equations in Table A-4 use a rate less than 85%. In the low- pressure quantity equations (7 and 15), the coefficient on GNP72$ goes from -5.53 (t: -O.933) to -7.79 (t: -1.71) when the cut-off rate is increased from 83% to 85%; that is, as we add Observations from time periods of somewhat higher capacity utilization rates the eXport quantity is more negatively affected (and the coefficient is significant at a higher level). In the comparable price equations (8 and 16), the coefficient on GNP72$ goes from 1.04 (t: 3.73) to 1.45 (t: 4.49); that is, GNP becomes more strongly correlated with the eXport price. Both patterns--price as well as quantity-- are 93 are consistent with the DPH. Support for the DPH, however, disappears when cut-Off rates of 86% and 87% are tried. A conclusion that is intuitively appealing and which is consistent with these estimations is that increases in domestic demand pressure affect eXport supply (price and quantity) of this commodity along the DPH lines; but that the effect is felt at relatively low capacity utilization rates as these rates increase. Once the capacity utilization rate reaches 86%, there is no longer any relationship between GNP and eXports. This might happen, for example, if this industry reaches full capacity before the rest of the economy does and, hesitant to raise prices even further, managers allocate output on a random basis or with some sense of "equity," Of sharing the shortage, or of loyalty to Old customers. This result would also be noticed if the industry had idle facilities that were "mothballedJ' Considering only the facilities currently being used, this industry might reach capacity output even though the economy-wide capacity utilization rate is less than 86% (and thus the demand pressure effect will be felt). As the capacity utilization rate rises above 86%, this industry might then decide to activate its idle, stand-by facilities. When this occurs, there is no longer a constraint on capacity, and there will be no evidence of a demand pressure effect. The start-up Of these additional faci. expOI demar Sampl subset as 30; A~6> 94 facilities could be used to satisfy the extra demand, and exports need not be affected by the increase in domestic demand. For the other two commodities from Section 5--533 2000, Printing inks (Tables A-5 and A-6), and 588 3060, Rubber cement (Tables A-7 and A-8)--there is support for the DPH in only one equation, that being the quantity equation of Model C for rubber cement; and there are three instances Of results that contradict the DPH. These contradictory results appear in the quantity equation of Model B for printing inks (Equation 1, Table A-5), where a GNP72$ coefficient of 6.24 (t: 2.75) is Significant at the 1% level, and in two price equations for rubber cement. These latter two equations are from Model E (Equation 6, Table A-7) and from Model C (Equation 10, Table A-8). In Model E, the coefficient on YHCAP87 is negative bn0287) and significant at the 10% level (t: -1.79). Because this is a price equation, a positive sign was eXpected. In Model C, the GNP72$ coefficient of -14.9 (t: -2.88) is significant at the 5% level. Once again, a positive coefficient was eXpected. The estimation results Of Model C, which divides the sample into a subsample with low demand pressure and a subsample with high pressure, can be interpreted cautiously as supporting the DPH in the case of printing inks (Table A-6). The price equations (8 and 10) have insignificant corre SUppo utili rate deman Fate commOI which Can nc Oxides SUDDOr 95 coefficients for GNP72$; but the quantity equation at low pressure (equation 7) has a GNP72$ coefficient of 6.21 with a t-ratio Of 2.18, which is significant at the 5% level (2-tailed test), and the quantity equation at high pressure has a GNP72$ coefficient of 3.49 with a t-statistic Of 0.626, which means that the 3.49 figure is not significantly different from zero at the 10% level or better (2-tailed test). Thus, as GNP rises, so do exports of printing inks--but this positive correlation disappears after domestic demand pressure increases enough to bring the capacity utilization rate to 87% or above. This transition from a positive correlation to zero correlation might be interpreted as weak support for the DPH. Of the four tested cut-Off rates for capacity utilization (83, 85, 86, and 87 Percent), the apprOpriate rate is 83% in the case Of one commodity (525 6030); domestic demand starts affecting eXports when the capacity utilization rate reaches 87%. Printing inks (533 2000) is one of the commodities that showed no support for the DPH no matter which cut-Off rate was used. The results of the tests of commodities from Section 5 can now be summarized. Three commodities were tested: iron oxides, printing inks, and rubber cement. Iron oxides support the DPH in six out Of fourteen tests. The support comes from both price and quantity, and it appears in Models E) su; DPI- in mo. SU Cc C< 96 B and C. There is one result (the quantity equation of Model E) that strongly contradicts the DPH. Printing inks Offer no support whatsoever for the DPH. Rubber cement supports the DPH in one quantity equation and strongly contradicts the DPH in two price equations. C. Commodities from SITC Section s (Manufactures). In the case Of thirteen Section 6 commodities tested, most tests do not support the DPH. As in the case Of SITC 5, support for the DPH appears more Often in the quantity equation (19 times) than in the price equation (13 times). A commodity from this section (691 11020--Door and window sash, frames, moulding and trim, of iron and steel) is one of the three that support the DPH more often than not (the other two are sewing machine needles and ball point pens, to be discussed below with SITC 7 and SITC 8, respectively). Also in this section are two commodities (aluminum wire and hacksaw blades) that do not support the DPH in any of the tests. Commodity NO. 641 1000, newsprint paper, gives conflicting results with respect to the capacity utilization rate at which domestic demand starts affecting eXports. Consider first the quantity equations in Tables A-9 through A-12. Equation 1 (Model B) Shows a GNP72$ coefficient of -147 (t: -3Jfl0, which supports the DPH. In Equation 3, which is an estimation of Model D, the GNP72$ term has been l‘f 26 (t tr. af YH co 50 ex on ad CO arI mu‘ PM the em (t: fro ind dif 97 replaced by YHCAP87 (which is equal to GNP72$ whenever the capacity utilization rate is 87 or higher and is equal to zero otherwise), and this term has a coefficient of -8.53 (t=-2u36). This is consistent with the DPH and indicates that a capacity utilization rate of 87% or greater adversely affects eXports. Equation 5 includes both GNP72$ and YHCAP87 (this is Model E) and the coefficients are consistent with the ones from Equations 1 and 3. Equation 5 suggests that increases in domestic real GNP restrict exports at all capacity utilization rates (the coefficient on GNP72$ is -148, with t = -3.87) but that there is an additional restriction when utilization is 87% or above (the coefficient on YHCAP87 is -9.13, with t = -2.78). However, Equations 9, 11, 13, and 15 suggest that the appropriate utilization rate is 83%. These four equations are estimations Of Model C, which divides the sample into two mutually exclusive sets based on capacity utilization rates. Equation 13 is the low-pressure and Equation 15 is the high- pressure version for a cut-Off rate Of 87%. The DPH predicts that the coefficients on GNP72$ for the high-pressure equation will be negative, but the estimated value is 63.4 (t: (L531). The coefficient is not significantly different from zero; indeed the F-statistic for Equation 15 (F = 05738) indicates that none of the coefficients are significantly different from zero in that equation. Equations 9 and 11 are test the coef pred is s the l is 83 and E resu] as Eq varia 11, t ShOul COeff: Signii COeffi the pr Statis test (. Fc paper; the def rate 0f 871.,”d definiti 98 tests of the same model, but here the criterion for dividing the sample is a capacity utilization rate Of 83%, and the coefficient on GNP72$ in the high-pressure version has the predicted negative Sign (-130) and, with a t-ratio of -1.68, is significant at the 5% level. These results suggest that the DPH holds and that the crucial capacity utilization rate is 83%. However, using Model E to distinguish between 83% and 87% the evidence is clearly in favor of 87%. The Model E results are presented for comparison purposes in Table A-10 as Equations 7 and 8. These equations use 83% for the shift variable YHCAP83. In order to be consistent with Equation 11, the coefficient on YHCAP83 in the quantity equation should be negative and significant. The estimated coefficient is positive (4.42) and not statistically significant at the 10% level. In the price equation, the coefficient is admittedly close to zero, but it does not have the predicted positive sign (it is -0.0154) and it is statistically significant at the 5% level in a two-tailed test (t = -2.15). Following is a summary of the results for newSprint paper: Model B supports the DPH; Model C supports the DPH if the definition of high pressure is a capacity utilization rate Of 83%-and-higher, but says nothing if the definition is 87%-and-higher; Models D and E support the DPH if the definition is 87%-and-higher; but Model B, in direct Opposit of high 83%-and It it comi Indeed, when tr definit result. What i: PESult Statis. faVor Here, , "behev Signif Inall partic 99 Opposition to Model C, contradicts the DPH if the definition of high demand pressure is a capacity utilization rate of 83%-and-higher. It is not inconsistent with the DPH to find support for it coming from more than one definition of high pressure. Indeed, a pattern like this is to be expected. Therefore, when there are consistent results with the various definitions, only one definition is used in the results reported here for each commodity. If the models support the DPH for a particular commodity but the different models imply two different definitions for high demand pressure, then the results for both definitions are reported in all models. What is inconsistent with one's eXpectations is a set of results such as those pertaining to newsprint paper. Statistically, either one rejects the null hypothesis (in favor of the DPH) or one cannot reject the null hypothesis. Here, we are claiming that the evidence supports the DPH whenever the pressure coefficient is statistically significant in a one-tailed test and has the predicted sign. In all other cases the evidence does not support the DPH, but particular attention is paid to the special cases where the pressure coefficient has the Opposite Sign from that predicted and is statistically significant at the 10% level or better in a two-tailed test. It seems reasonable to interpret results of this nature as strong evidence contradicting the Demand Pressure Hypothesis. 100 The test results for the commodities from Section 6 can now be summarized. Thirteen commodities were tested: newsprint paper, pig iron, concrete reinforcing bars, cOpper wire, COpper powder, aluminum wire, aluminum powder, zinc sheets, steel door frames, hacksaw blades, twist drills, and razor blades. Two of these, aluminum wire and hacksaw blades, Offer no support for the DPH. Three commodities--pig iron, COpper wire, and aluminum door frameS--support the DPH in at least one quantity equation but not in any price equations. COpper powder, aluminum powder, and razor blades support the DPH in at least one price equation each but not in any quantity equations. The remaining five commodities support the DPH in at least one quantity equation and at least one price equation. Five commodities strongly contradict the DPH in one equation each, and three do so in two equations each. D. Commodities from SITC Section 1 (Machinery). Among the Section 7 commodities the following three give ineXplicably inconsistent results in the quantity equations: 742 4026, centrifugal pumps; 751 1040, electric typewriters; and 775 7520, vacuum cleaners. For centrifugal pumps the results are shown in Tables A-50, A-51, and A-52. The inconsistency is between Equation 3 and Equations 11, 12, 13, and 14, all of which are quantity equations. Equation 3, which is Model D, suggests that the quantity of eXports is adve capac 0n YE diffe These by ti Here, signi Furtr using MOdel YHCAP the 1 C (Eq Quant and (T1 equat °” thI 101 adversely affected by domestic demand pressure when the capacity utilization rate is 83% or higher. The coefficient on YHCAP83 is -7.09 and, with a t-ratio Of -1.32, is significantly less than zero at the 10% level. However, if the same model is run with a capacity utilization rate of 85% (Equation 11), the coefficient on YHCAP85 is 8.42 with a t-statistic Of 1.47, which is significantly greater than zero in a one-tailed test at the 10% level and not significantly different from zero in a two-tailed test at the 10% level. These results, which do not support the DPH, are reinforced by the estimation of Model E using YHCAP85 (Equation 12). Here, the coefficient is again positive (16u8) and is significant (two-tailed test) at the 1% level (t = 2(T7L Further evidence strongly contradicting the DPH is found by using YHCAP86 in the estimation of Model 0 (Equation 13) and Model E (Equation 14). In both cases the coefficient on YHCAP86 is positive and significantly different from zero at the 1% level. For electric typewriters, the quantity equation of Model C (Equation 9 in Table A-57) supports the DPH; but the quantity equations (and also the price equations) of Models D and E (Table A-56) strongly contradict the DPH. ‘The Model B equations do not have statistically significant coefficients on the pressure variable. For vacuum cleaners, the inconsistency lies in the reverse direction from that of centrifugal pumps: here, the evic the but regr eQua equa has zero 10) Show Pric nega 8th Meto 102 evidence contradicts the DPH if 83%-and-higher in chosen for the capacity utilization rate, but supports the DPH at a rate of 85% or higher (Tables A-62, A-63, A-64, and A-65). When the capacity utilization rate is at least 85%, Equations 3 and 5 (Models D and E) indicate that the quantity of eXports falls as GNP rises, but Equation 9 (Model C) indicates a positive correlation between GNP and the quantity of eXports. If high pressure is defined as a capacity utilization rate of 83%-or-higher, the quantity equation (17) Of Model C and the price equations (12 and 14) of Models D and E contradict the DPH. With this definition, there is no support for the DPH. One commodity--762 0040, Radios, household type, without phonograph--shows support for the DPH in a quantity equation but contradicts the hypothesis in a price equation. The regression results are in Tables A-58 and A-59. The only equation supporting the DPH is NO. 3, which is the quantity equation of Model D, where the YHCAP87 coefficient of -22.6 has the eXpected negative sign and is significantly less than zero at the 5% level.(t: -2.29). 'The price equations (8 and 10) of Model C contradict the DPH. ‘The low-pressure equation shows a significant positive correlation between GNP and price, while the high-pressure equation shows a significant negative correlation--the DPH predicts a positive correlation at high pressure. An interesting feature of Commodity NO. 716 4042-- Motors, AC, polyphase, induction, not over 20 hp--is that the rate press than COeff; Signij in eit Model coeffi models greatE Th1s c domest affPCt. rates, pres‘SUr reSPEQt 103 rate of capacity utilization at which domestic demand pressure begins to affect the quantity Of exports is lower than the rate at which demand pressure begins to affect the price Of the eXports (Tables A-39 through A-42). Using a capacity utilization cut-Off rate Of 85%, the quantity equations for Model D (Equation 3) and Model E (Equation 5), Show the expected negative coefficient on YHCAP85 and each is significant at the 1% level; but in the price equation, the YHCAP85 coefficient is not Significant in Model D and is very small (.0138) and significant at only the 10% level in Model E. When the cut-off rate is changed to 87%, the coefficients on YHCAP87 (which replaced YHCAP85) are not significantly different from zero in the quantity equations in either Model (Equation 11 is Model D and Equation 13 is Model E). In the price equations (12 and 14), the coefficients on YHCAP87 are virtually identical in the two models (0.0367 in D and CL0368 in E), and are significantly greater than zero at the 1% level in a one-tailed t-test. This commodity supports the Demand Pressure Hypothesis. As domestic demand increases, the quantity of eXports is affected first. Only later, at higher capacity utilization rates, is there pressure on prices to start rising. Combines (NO. 721 2220) behave according to the Demand Pressure Hypothesis with respect to price but not with reSpect to quantity (Tables A-43, A-44, and A-45). The price not some Tabl equa defi less samp high incr 1.41' Of” equa 104 equations are Equations 2, 4, and 6, which are from Models B, D, and E, respectively. In all three cases, the pressure coefficient has the expected positive Sign and is significant at the 1% level. Model C (Equations 7, 8, 9, and 10) does not support the DPH, but an adaptation of Model C provides some interesting results. Equations 11, 12, 13, and 14 in Table A-45 are the low--pressure versions of the price equations Of Model C as the capacity utilization rate in the definition of low pressure is changed from less than 83%, to less than 85%, to less than 86%, to less than 87%. As the sample is enlarged to include observations from periods with higher capacity utilization rates, the coefficient on GNP72$ increases monotonically from -2.72 to 7.58 to 11.8 to 17.7. The t-ratio also rises montonically: -0.149, 08709, 1.41*, 2.31‘*. Thus, even though the high-pressure equations of Model C do not support the DPH, the low-pressure equations do Offer support as the definition of low pressure is broadened to include higher demand pressure. It appears that increases in domestic demand pressure lead to higher eXport prices, but only up to a certain point. Once the economy-wide capacity utilization rate reaches 87%, the eXport prices in this industry no longer have any connection with the domestic economy. One eXplanation might be that the eXport price is initially lower than the world price Of the closest substitutes. As domestic demand increases, the expo pric furt hous lend this coefl signi bette and t 105 eXport price rises until it is equal to the substitute's price, at which point the forces Of competition prevent any further price increase. Commodity No. 775 8625--Toasters, automatic, electric, household type--is the one commodity from SITC Section 7 the lends no support whatsoever for the DPH. Four equations for this commodity (Tables A-66 and A-67) have pressure coefficients with the uneXpected sign and yet are significantly different from zero at the 10% level or better. These are the quantity equations of Models B and C and the price equations of Models D and E. The test results for the commodities from Section 7 can summarized as follows. Eleven commodities were tested: electric motors, combines, dozers, sewing machine needles, centrifugal pumps, air compressors, electric typewriters, radios, electric shavers, vacuum cleaners, and toasters. Only one commodity--toasters--Offered no support whatsoever for the DPH. Three commodities--typewriters, radios, and vacuum cleaners--supported the DPH in at least one quantity equation but not in any price equation. Combines did just the reverse, Offering support in (four) price equations but not in any quantity equations. The remaining Six commodities supported the hypothesis in both price and quantity equations. Five commodities--centrifugal pumps, typewriters, radios, vacuum cleaners, and toasters--showed coefficients that strongly contradicted the DPH. A~7I> when 16), YHCAP QOeff indic incre rise; coeff signi; 106 E. Commodities from SITC Section 8 (Miscellaneous Manufactures). Among the four commodities tested from Section 8, two (884 2220, sun glasses, and 891 0945, plastic tape) exhibit inconsistencies in the price equations, with plastic tape offering support for the DPH in the quantity equation and the other not Offering any support in the quantity equation. A third commodity (885 2020, electric clocks) supports the DPH in two price equations but strongly contradicts the hypothesis in one of the quantity equations. The fourth commodity (895 2115, ball point pens) supports the DPH both in price equations and in quantity equations at the 10% level or better and has no equations that Show significant pressure coefficients with an unexpected Sign. In the regressions for sun glasses (Tables A-68 through A-71), there is support for the DPH in the price equations when the shift variable YHCAP85 is used (Equations 4, 6, and 16), but the results are contradictory to the DPH when YHCAP86 is used instead (Equations 12, 14, and 18). The coefficients on YHCAP85 are significantly greater than zero, indicating that at high levels of demand pressure any increase in real domestic GNP will cause export prices to rise; this is, Of course, consistent with the DPH. The coefficients on YHCAP86, however, are negative (and significantly different from zero), suggesting just the Opposi 0f the 107 opposite of what the DPH predicts. For this commodity none of the quantity equations have pressure coefficients significantly different from zero. The price equations for plastic tape (Tables A-75 and A-76) support the DPH in Models D and E (Equations 4 and 6) when the dividing line between low and high pressure is a capacity utilization rate Of 86%. The coefficients on YHCAP86 are 0.134 (t = 1.92) and 0.132 (t: 1.90) for Model 0 and Model E, respectively, and both of these are significantly greater than zero at the 5% level. But using the same dividing line (86%) in Model C, which is given as Equations 8 (low pressure) and 10 (high pressure), the coefficient on GNP72$ in the high-pressure equation is negative (-57£fl rather than the predicted positive and (with t = -3.06) is significantly different from zero at the 1% level in a two-tailed test. With the quantity equations, Model B contradicts the DPH, Model C supports it, and the null hypothesis cannot be rejected in the other models. The results for electric clocks (885 2020) are similar to those Of sun glasses: the quantity equations Offer no Obvious support for the DPH (Tables A-72 through A-74). Indeed, the quantity equations of Model C indicate that at low demand pressure an increase in GNP will adversely affect the quantity of eXports but at high demand pressure there is no relationship between the two variables. If YHCAP85 is used stron insig is us DPH. oxide may r does. low I dome: OConr has 1 disc: no s EXDQ 108 used as the shift variable in Model E, the quantity equation strongly contradicts the DPH and the price equation has an insignificant coefficient on the YHCAP85 term. If YHCAP83 is used, however, the price equation of Model D supports the DPH. However, the conclusion suggested earlier for iron oxides (525 6030) might be appropriate here: this industry may reach capacity output before the rest of the economy does. While the rest of the economy is still experiencing low demand pressure, this industry is reducing eXports as domestic sales increase. By the time the rest of the economy is eXperiencing high demand pressure, this industry has reached capacity output but does not systematically discriminate against the foreign purchaser; and, therefore, no significant relationship shows up between industry exports and aggregate demand pressure. Commodity NO. 895 2115--Pens, ball-point type--supports the DPH in both the price equations and the quantity equations when high demand pressure is defined as a capacity utilization rate Of 83% or higher (Tables A-77 and A-78). The quantity equations (3, 5, and 9) Of Models 0, E, and C show up with predicted (and significant) sign on the pressure variables. As for the price equations, support comes from Model C but not from Models B, D, or E. If a cut-off rate of 85% is used, support is lost from the quantity equations of Models D and E (Equations 11 and 13) but equa' appe and now glas All doin ball Pric thre Cont Simi equa equa the Cont The Ofte Pres Vari resu 109 but is maintained in Model C (Equation 17). For the price equations, support continues in Model C (Equation 18)and appears for the first time in Models D and E (Equations 12 and 14). The test results for the commodities from Section 8 can now be summarized. Four commodities were tested: sun glasses, electric clocks, plastic tape, and ball-point pens. All four support the DPH, sun glasses and electric clocks doing so in price equations only, while plastic tape and ball-point pens offer support in the quantity as well as the price equations. Although sun glasses support the DPH in three price equations, paradoxically they also strongly contradict the DPH in two different price equations. Similarly, plastic tape supports the hypothesis in two price equations and strongly contradicts it in another price equation. With respect to electric clocks, the support for the DPH comes from the price equations, and the strong contradiction of the DPH shows up in a quantity equation. The fourth commodity, ball-point pens, supports the DPH more often than does any other commodity tested. F. The Tables. The tables containing the regression results are presented as an appendix below in numerical sequence by commodity number. Figure 5-2 describes the independent variables and the other notation used in the tables Of results. Capac Const. VAROW XRIMF 0:2723 PL PK D781791 DUM YHCAP Rho R‘SQUam H df(num,d. 0w, FiEUre 5. Capacity Util. Constant VAROW XRIMF GNP72$ PL PK 0781794 DUM YHCAP__ Rho R-Squared F: df(num,den) DW = Figure 5-2. 110 This pertains to Model C only and indicates the capacity utilization rate that was used to divide the sample into low-pressure and high-pressure subsamples. .LT. = Less than; .GE. = Greater than or equal to. For example, ".LT.83" means Capacity Utilization is less than 83%. The constant term in the regression. (For all variables, the numbers in parantheses are t-ratios) Value Added, Rest of WOrld (World except U.S.). Exchange rate index from IMF'S Multi-lateral Exchange Rate Model. The value of the Dollar rises when XRIMF rises. U. S. GNP in 1972 dollars. Unit labor cost, non-farming, deflated. Unit non-labor cost, non-farming, deflated. This is a proxy for the price of land and the price of capital combined. Dummy variable. Equals 1 from 1978.1 to 1979J1 and equals 0 otherwise. Dummy variable. For example, DUM653 equals 1 during 1965.3 and equals 0 otherwise. Interaction term between Real Home Income and high capacity utilization rates. For example, YHCAP83 equals UAR GNP in 1972 dollars when the capacity utilization rate is 83% or higher and equals zero otherwise. The auto-regressive RHO. From U(t) = RHO * U(t-1) + e(t). Coefficient of determination. F-Statistics. The degrees Of freedom for the numerator and for the denominator Of the F-Statistic. Durbin Watson Statistic. EXplanation Of Variable Names. resu Dema for incr eXpo quan incn POStI modei for 1 when Cochr 9“ th Unite Quart CHAPTER 6 CONCLUSION This Chapter summarizes the study and shows how its results fit into the current body of literature on the Demand Pressure Hypothesis. In addition, some possibilities for future investigation are suggested. I. Summary. The Demand Pressure Hypothesis postulates that increases in domestic demand would have an adverse impact on exports. Adverse impact is interpreted to mean that the quantity will decrease and/or the export price will increase. Starting with traditional microeconomic postulates about firm behavior, four different testable models were develOped. The reduced forms for quantity and for price were estimated with Ordinary Least Squares and, when necessary, with an autoregressive model using a Cochrane-Orcutt type iterative procedure.19 Tests were run on thirty-one seven-digit commodities eXported from the United States for the period 1965.1 through 1979.4 using quarterly observations. This study departs from previous studies in several ways. First, it retains the simultaneity of both price and quantity by using reduced forms rather than a single 19The regression package was Version 2.4 of SHAZAM (see White, 1978). equ rig Phy gal dis def con lev aha funI OVer COmr eij beir 0fE uSir wher indL dete menu Sens As rish cost, 112 equation model with quantity on the left and price on the right Side. Second, the quantity figures are actual physical units such as metric tons, number of items, or gallons. This is made possible by a high level Of disaggregation. Third, the commodities are more narrowly defined, being at the seven-digit level rather than the conventional four-digit level or even the "all manufactures" level. Fourth, factor costs are brought explicitly into the analysis; given that the DPH is a theory about a supply function, the previous neglect of input prices needed to be overcome. The same macroeconomic variables were used for all commodities. Although a case can be made for using eXplanatory variables that are Specific to the commodity being tested, the use of macro variables avoids the problems of simultaneity. For example, one would not be comfortable using a capacity utilization rate for a seven-digit industry when trying to explain the quantity of eXports Of that industry, because the two variables are simultaneously determined. The capacity utilization rate for all manufacturing industries combined, however, should not be sensitive to the eXport success of one seven-digit industry. As another example, consider the price equations and the two right-hand variables unit labor cost and unit non-labor cost. If those two explanatory variables come directly from 113 the same seven-digit industry as does the eXport price, for all practical purposes the estimation procedure is being applied not to an equation but to a definition. The capacity utilization rate for total manufacturing is used as the indicator of domestic demand pressure. This variable does not enter the regressions directly. Instead, it is used to divide the sample into mutually exclusive subsets of high demand pressure and low demand pressure. Five models were constructed. Model A is the null hypothesis and asserts that eXports are unrelated to domestic demand. Model B claims that eXport quantity and price are related to domestic demand at all times. Domestic demand is measured by Gross National Product in 1972 dollars. Model C asserts that the coefficients of the explanatory variables are different during periods Of high domestic demand pressure than they are during periods of low domestic demand pressure. Using the capacity utilization rate to separate the sample into periods of high demand pressure and periods of low demand pressure, the equations of Model C were estimated separately for the two subsamples. Model D claims that the relationship between eXports and all the eXplanatory variables except real Gross National Product is the same regardless Of the capacity utilization rate. With respect to GNP, the model asSerts that changes in G rate high that is h clai when numb incl on G Meth inte capa 0the Fela YHCA Peri 114 in GNP do not affect eXports if the capacity utilization rate is low, but they do if the capacity utilization rate is high. The procedure employed is to include a term (YHCAP) that is equal to GNP72$ when the capacity utilization rate is high and is equal to zero otherwise. Model E is one step more general than Model D. Model E claims that the relationship between GNP and eXports changes when the capacity utilization rate reaches some critical number indicating high pressure on capacity. This Model includes GNP72$ in the equation, but allows the coefficient on GNP72$ to change at high capacity utilization rates. Methodologically, this is accomplished by the addition of an interactive term (YHCAP) that is equal to GNP72$ when the capacity utilization rate is high and is equal to zero otherwise. The coefficient of GNP72$ indicates the average relationship between eXports and GNP, and the coefficient of YHCAP indicates the change in this relationship during periods of high domestic demand pressure. The data for twenty-seven of the thirty-one commodities lent at least some support to the Demand Pressure Hypothesis. As a summary statistic, the Hypothesis was supported in roughly one-third Of the tests--specifically, in 89 out of 305 tests. The Hypothesis was contradicted in 26 quantity equations and 16 price equations; that is, the apprOpriate coefficient was statistically significant at the 10% pred them each easi thre thir seve inst. 115 10% level or better, but had a Sign Opposite of that predicted by the DPH. II. Comparative Performance pl the Models. Also of interest is the performance of the models themselves. Considering the number Of instances in which each model supports the DPH, Models C, D, and E cannot easily be distinguished from each other, but each of these three is superior to Model B. Model B supports the DPH thirteen times, Model C twenty-four times, Model D twenty- seven times, and Model E twenty-three times. In every instance in which a quantity equation from Model B supports the DPH, at least one quantity equation from the other models also supports the hypothesis. With two exceptions (sewing machine needles and centrifugal pumps) an analogous statement for price equations is also true. The feature that distinguishes Model B from Models C, D, and E is that Model B asserts that increases in domestic demand have adverse effects on export performance at all levels Of domestic demand pressure, whereas the other three models limit the adverse effects to periods of high demand pressure. Thus, the hypothesis underlying Models C, D, and E is clearly preferred to that of Model B. Domestic demand is more likely to affect eXports when the capacity utilization rate is already high. Recall that the Average Total Cost curve in Model B is U-shaped and in Models C, D, and cape for DPH Modl Of L (inc SUE. con hor thi nur Ut‘. T11 DP Su (A LJU’ 116 and E has a horizontal range and then lepes upward at capacity output. Furthermore, if ATC remained horizontal for a particular commodity throughout the sample period, the DPH would not be supported. Thus, support for the DPH in Model B would be the only evidence in favor of the existence of U-shaped Average Total Cost curves. All other cases (including the 216 tests that did not support the DPH) suggest that ATC curves have a horizontal portion. The conclusion that Average Total Cost curves are indeed horizontal over a range seems an inescapable implication of this study.20 III. Comparative Performance sf Cut-Off Rates. For each of the three models (C, D, and E) requiring a numerical definition Of full-capacity, four capacity utilization rates were tested: 83, 85, 86, and 87 percent. The cut-Off rate that is most successful in supporting the DPH is 86 percent. The number Of times that each rate supported the DPH is as follows: seventeen times for 83%, sixteen times for 85%, twenty-three times for 86%, and eighteen times for 87%. The value of testing four different capacity utilization rates for each commodity is that this technique allows demand pressure to be experienced in one 20This is consistent with several of the more-direct tests of cost functions. A good introduction to the literature on statistical cost estimation can be found in Johnston (1960) or in Dean (1976). ind ind can 0th Of I gre; SOI'G eCOI dome Comr Cla: bott the Cap; adve 117 industry before it becomes a constraint in some other industry. Thus, when the economy is eXpandihg, bottlenecks can start appearing in some industries sooner than in others, and this technique uncovers that information. lpslssl one would eXpect that the greater the number of manufacturing stages the commodity goes through, the greater the Opportunities for bottlenecks to appear somewhere in that process, and thus, the lower will be the economy-wide capacity utilization rate at which increases in domestic demand will adversely affect exports of the commodity. Conversely, the more nearly the item can be classified as a raw material, the fewer the chances for bottlenecks to interrupt the production of that commodity as the economy eXpands, and thus, the higher the economy-wide capacity utilization rate at which domestic demand pressure adversely affects exports. Broadly speaking, the evidence supports this expectation. Figure 6-1 classifies the support for the DPH according to the capacity utilization rate used in Models C, D, and E. It is difficult to draw any conclusions regarding the commodities from Section 5 because of the limited support for the DPH in these Models. But the overall results for Sections 6 and 8 reflect the effect Of the stage of production alluded to above. The commodities in Section 6 are primarily industrial materials that will undergo further proc of t thes defi occu Sect for defi elex thre Comm Cate are and ; will To U util the , One c commC affec °apao 118 processing or be used as tools to produce final items. Out Of the twenty-eight instances of support for the DPH from these commodities, twenty-four occur when full capacity is defined as 86%-and-higher or 87%-and-higher and only four occur at lower definitions. In contrast, the commodities in Section 8 are primarily final consumer items. Here, support for the DPH is concentrated at the lower values for definitions of full capacity. Support for the DPH occurs eleven times with cut-Off rates Of 83% and 85%, and only three times with rates Of 86% and 87%. Both groups of commodities behave according to eXpectations. The commodities drawn from Section 7 are not as easy to categorize. Some, such as electric shavers and toasters, are final household items. Others, such as electric motors and self-propelled combines, are final business items that will be used in the production of other goods and services. To the extent that they are all final items, low capacity utilization cut-off rates are eXpected to be successful in the regression models, but no such behavior is apparent. One discernible pattern that does appear is that for these commodities as a group, the export gpantity begins to be affected adversely by domestic demand at lower rates of capacity utilization than does the eXport pric . Comp Numb 525‘“ 533 . 588 641 671 673 682 682 684 684 686 ; 691 ‘ 691 ; 695 g 695 I 696 ( AOA¢-Ay~—‘t 119 Capacity Utilization Rate Commodity QX PX Number Brief Description 83 85 86 87 83 85 86 87 525 6030 Iron oxides 1 1 1 1 533 2000 Printing inks 588 3060 Rubber cement 1 "777701 1777570 641 1000 Newsprint 1 2 1 671 2000 Pig iron 1 673 2005 Concr.reinforc.bars 1 1 682 2160 COpper wire 2 682 2400 COpper powder 1 684 2140 Aluminum wire 684 2420 Aluminum powder 1 686 3220 Zinc sheets 3 1 691 1020 Steel door frames 3 3 691 2020 Aluminum door frames 2 695 3140 Hacksaw blades 695 4145 Twist drills 1 2 696 0340 Razor blades 2 Figure 6-1. Comparison Of Capacity Utilization Rates. (Continued on next page.) Comp Numb 716 721 723 724 742 743 751 762 775 775 ' 775 . 884 885 891 895 (\\r\h\ 1“ ~ ---- 120 Capacity Utilization Rate Commodity QX PX Number Brief Description 83 85 86 87 83 85 86 87 716 4042 AC motors ‘2 1 “E 721 2220 Combines 2 723 4052 Dozers 1 1 724 3920 Sewing mach. needles 3 742 4026 Centrifugal pumps 1 743 1035 Air compressors 2 1 3 751 1040 Electric typewriters 1 762 0040 Radios 1 775 4030 Electric shavers 2 2 775 7520 Vacuum cleaners 2 775 8625 Toasters 884 2220 Sun glasses 2 885 2020 Electric clocks 1 891 0945 Plastic tape 1 2 895 2115 Ballpoint pens 3 1 1 3 ‘3‘1‘1’0’ “2'5 2‘0 NOTE: Entries denote the number of times that an equation using that capacity utilization rate for the definition Of high pressure supported the Demand Pressure Hypothesis. The OX columns refer to the quantity equations and the PX columns refer to the price equations. The relevant Models are C, D, and E (see Figure 5-1). Figure 6-1 (cont.) call peri dema capa Of e duri used quan' restI Vari for ' be bC equal high this the I and e alreE (31th only infOr commo r9801 121 IV. Henry's Weak Test. Chapter 2 contains a description of what Henry (1970) calls a weak test of the DPH. It postulates that during periods of high domestic demand, the link between world demand and eXports is broken. That is, when there is excess capacity at home, world demand is a significant determinant Of eXports, but demand for these eXportS remains unsatisfied during periods of high pressure on capacity. Model C can be used to test this hypothesis. Henry considered only the quantity of eXports, so the following test will be restricted to the quantity equations. The world demand variable is Value Added Rest of World (VAROW). The criteria for "passing" the test will be that the coefficient on VAROW be both positive and significant in the low-pressure equation but not positive and significant in the high pressure equation. Four out of the 31 commodities pass this test and, under Henry's interpretation, would support the DPH: newsprint paper, air compressors, vacuum cleaners, and electric clocks. However, each of these commodities already supports the DPH without resort to this weak test (although electric clocks support the DPH in price equations only and therefore the test does provide some supplementary information). What would have been helpful is if the four commodities lending absolutely no support for the DPH in the regular test had been able to pass this weak test. rea con mod (19' exi dom. eXp: Prol Phi. the M Opt: pat! the 122 V. Exports ls s Residual. Next to be examined is the eXports-are-a-residual argument. It maintains that manufacturers will keep their plants running at full capacity, sell what they can at reasonable price at home, and ship the left-over production to the export market for sale at any price obtainable. The argument has a certain amount Of appeal, but it lacks firm analytical foundations. It requires, among other things, that full capacity output be synonymous with Optimum output. Alternatively, one could assert that entrepreneurs are not Optimizers, in which case economic analysis has little to contribute. NO one has developed a satisfactory economic model which implies that eXports are a residual. Henry's (1970) model comes the closest, but it relies on the existence of excess demand and a rationing rule that favors domestic customers. There also is the implication that the eXport price is irrelevant; that the firm has made its production decision (how much to produce) without regard to price (or marginal revenue). There exists a model which leads to a special case Of the exports-are-a-residual view; but even there the word residual is misleading because it turns out that the Optimal quantity Of eXports varies in a one-to-one Offset pattern with the quantity sold at home. The reference is to the dumping model with a downward SIOping domestic demand curve upwar 13). as it equa' dete deMa tote at w PFOI exi, rev the the (fie th QC 0 Ho 123 curve, an infinitely elastic foreign demand curve, and an upward leping marginal cost curve (see Figure 2-1 on page 13). In such a model, the eXport price cannot be dismissed as irrelevant because it is the eXport price (which is equal to the marginal revenue from the eXport market) that determines the Optimum quantity to produce. The domestic demand and marginal revenue curves determine the portion of total output that is sold, in the home market and the price at which this output that is sold, but the total quantity produced depends on the export price. This dependence exists because the horizontal summation of the two marginal revenue curves will yield a combined marginal revenue curve that becomes a horizontal line at the same dollar value as the export price. Ruling out the case where an increase in domestic demand is so great that eXports fall to zero, it is clear that as domestic demand increases, total production remains constant, so that for every additional unit sold at home one less unit is sold in the eXport market. Total production remains constant because if the demand for eXports does not change, the combined marginal revenue cannot change. The intersection of the marginal cost curve and the combined marginal revenue curve occurs at the same output as before. The interesting feature of this model is that the change in domestic demand does not affect the eXport price. A test Of the model would be that changes in domestic demand 124 are inversely related to the quantity of exports but are not correlated with the price of eXports. This hypothesis would be supported by the following evidence. In Model B, the coefficient on GNP72$ should be significantly less than zero in the quantity equation and not be significantly different from zero in the price equation. In Model C the coefficient on GNP72$ should be significantly less than zero in the high-pressure quantity equation and not be significantly different from zero in the high-pressure price equation. In Models D and E, the coefficient on YHCAP Should be significantly less than zero in the quantity equation and not significantly different from zero in the price equation. Of the thirty-one commodities tested, fifteen support this hypothesis in at least one Model each. Six of these commodities are from SITC Section 6, eight are from Section 7, and one is from Section 8. None of the commodities drawn from SITC Section 5 support the hypothesis. The commodities supporting the hypothesis are listed in Figure 6-2. VI. Evaluation 9: DPH. The Demand Pressure Hypothesis is now well-grounded in generally accepted economic theories Of producer behavior. What began as a proposition in macroeconomics, motivated by an interest in aggregate eXports and the trade balance, has evolved quite properly into a microeconomic theory; and it 125 is at the micro level that the theory must be tested. Given the intuitive appeal of the Hypothesis, discarding it would be difficult even if empirical research showed a total lack of support. Rather, the more attractive alternative would be the devising of increasingly more sophisticated tests. But there is empirical support for the Hypothesis, and this support is quite compelling. As the theoretical treatment of Chapter 3 makes clear, the shape of the Average Total Cost curve is a crucial factor in the relationship between eXports and domestic demand. In those instances in which the evidence does not support the DPH, it is tempting to conclude that either the tests or the data are not refined enough to detect the point at which the Average Total Cost curve starts sloping upwards. VII. Future Research. This research could be extended in several directions. First, it would be interesting to find out if micro-level explanatory variables are more useful than the macro variables used here. For example, following Artus (1970) one might be able to construct a time series for a capacity utilization rate for each individual industry. Problems of endogeneity would preclude the use of commodity-level capacity utilization rates, but industry-level rates might be helpful. Second, different mathematical specifications of t rele dist econ foun impr leas woul Comm 126 of the equations could be tried. For example, logarithmic relationships and various lag structures, such as polynomial distributed lags, could be hypothesized. Third, different econometric techniques could be employed. Although Dunlevy found that two-stage least squares offered no statistical improvement over single-equation Cochrane-Orcutt interative least squares, his tests were for total eXports; and it would be interesting to use the 2SLS procedure at the commodity level. 127 Evidence Supportigg the Hypothesis Commodity NO. Brief Description 641 1000 NewSprint 671 2000 Pig iron 673 2005 Concrete reinforcing bars 682 2160 COpper wire 686 3220 Zinc sheets 691 2020 Aluminum door frames 716 4042 AC motors 723 4052 Dozers 724 3920 Sewing machine needles 742 4026 Centrifugal pumps 743 1035 Air compressors 751 1040 Electric typewriters 762 0040 Radios 775 7520 Vacuum cleaners 895 2115 Ball-point pens TOE§I§?7“T§-Edfimddi€{€§ 7777777777777 Figure 6-2. The EXports-Are-A-Residual Model £4.21 D E B C B C D E D E D E D B C D E D D E C D D E D E 3 4 11 8 Hypothesis. APPENDICES APPENDIX 1 exchar value perio the n IMF' the of 1 the 0th did Cur Agr fir to .Ve: Calculation of XRIMF As mentioned in Chapter 4, the IMF'S index of effective exchange rate for the U.S. dollar begins in 1972.1; and so values for this index have to be extrapolated for the sample period from 1965.1 through 1971.4. This appendix describes the method used for constructing an index (called XRIMF) that measures the exchange value Of the U.S. dollar. The index XRIMF is simply a backwards extension of the IMF's index. This was accomplished by taking into account the devaluations and revaluations of the currencies fO four of the world's major countries--Canada, Germany, France, and the United Kingdom. The par values of the currencies of two other major trading partners of the UJL--Japan and Italy-- did not change during this period. It is assumed that the currency realignments arising out Of the Smithsonian Agreement of December 18, 1971, went into effect during the first quarter of 1972. The weights used are the dollar values of UAL eXports to that country as a percent Of total U.S. eXports for the year; the weights, therefore, change each year. Unfortunately, the weighting scheme differs from that of the IMF, so that some inconsistencies are introduced into the index. If a bias has been introduced, it is in the form Of too small a variation in the index values for the relevant 128 129 time periods. In absolute value, the largest adjustment is 1.8486, which is not a very large adjustment to an index that is equal to 100.0 in the base year, absence of MERM- derived weights for the period 1965.1 through 1971.4. Given the index for 1965.1 - 1971.4 takes on a constant value equal to the Fund-reported value for 1972JL The devaluations and revaluations that occurred and their effect on XRIMF are described below.21 On November 24, 1968, the Federal Republic of Germany lowered eXport subsidies from 11% to 7% and lowered border import taxes from 11% to 7%. Although this action affected only commodities, not services or capital, it is treated here as equivalent to a 4% revaluation. The reverse adjustments to the effective exchange rate index for the years 1968, 1967, 1966, and 1965 are, respectively, 0.19736, 0.21642, 0.22079, and 0.24021. The French franc was devalued 11 percent on August 8, 1969. Therefore, in the beginning of 1969 the dollar was worth less than in the final quarter of 1969. The adjustment from this source for the first three quarters of 1969 is -0.3459; for 1968, 1967, 1966, and 1965, the adjustments are -0.34776, -0.35746, -O.36536, and -0.38871 21For a description of these events in their historical context see Kreinin (1975, pp. 150-155). For a detailed account of the foreign exchange and balance Of payments history of these six countries plus the U.S. since World War II, see Yeager (1976, pp. 459-588). respec O percen 11 per action adjust: and No Quarte. for 194 TI Novemb -0.4441 Other 1 and for OI float , calCul. of the as thou Th1s me fixed 0 Of each 1971.4, 7L3228. Of a £0] 130 respectively. On October 24, 1969, the German mark underwent a 9.29 percent revaluation and the German government restored the 11 percent eXport subsidy and import border tax. This action is treated as 5.29 percent revaluation. The adjustment arising from these two actions (Oct. 24, 1969, and Nov. 24, 1968) combined is 0.29816 for the first three quarters of 1969 and 0.45836, 0.50263, 0.5128, and 0.55788 for 1968, 1967, and 1965 respectively. The 14.29 percent devaluation of the pound sterling on November 18, 1967, gives rise to an adjustment factor Of -0.44412 for the fourth quarter of 1967 and -O.88824 for the other three quarters. For 1966 the adjustment is -0.81881 and for 1965 it is -O.8403. On June 1, 1970, the Canadian dollar was allowed to float after having been pegged at C$1 = US$0.925. To calculate an adjustment factor, the exchange rate at the end Of the quarter (OECD, Main Economic Indicators) is treated as though it were the exchange rate all through the quarter. This method treats the Canadian dollar as though it were fixed but undergoing a devaluation or revaluation at the end of each quarter. For the six quarters from 1970.3 through 1971.4, the adjustments are -1.29182, -1.45104, -1.7016, -1.32287, -1.6759, and -1.8486. In this case the revaluation of a foreign currency causes a negative adjustment in the in Ii in 00 OC FE re 131 index because the revaluation occurs after the base period (May 1970 = 100) for the index. The revaluation reduces the index from 100. In the previous cases, revaluations occurred prior to the base period and the adjustments had to occur backwards. That meant that entries prior to the revaluation had to be increased so that at the time of the revaluation the index number declined towards 100. The Canadian case is the last adjustment to be made. The resulting index of effective exchange rate is assigned the name XRIMF. APPENDIX 2 F eight Stati' unit manhc manhc sect Calculation of PL and PK For measuring the prices of the factors of production, eight different series (1967 = 100) from the Bureau of Labor Statistics, Employment and Earnings, were tried: (1) W1, unit labor cost, manufacturing; (2) W2, compensation per manhour, manufacturing; (3) W3, real compensation per manhour, manufacturing; (4) W4, unit labor cost, private sector, non-farm; (5) W5, compensation per manhour, private sector, non-farm; (6) W6, real compensation per manhour, private sector, non-farm; (7) W7, unit non-labor payments, private sector; and (8) W8, unit non-labor payments, private sector, non-farm. Non-labor payments include profits, depreciation, interest, rental income, and indirect taxes. Real compensation per manhour, manufacturing, is merely compensation per manhour, manufacturing, divided by the Consumer Price Index (W3 = W2/CPI). Real compensation per manhour, private sector, non-farm, is Obtained in a Similar fashion: W6 = W5/CPI. W1, W2, and W3 are available for the entire sample period (1965.1 - 1979J0u For W4, W5, and W6 the published data begin in 1966; and for W7 and W8 the data begin in 1967. Values for the missing data were extrapolated in the following manner: W4 through W8 were each regressed against W1, W2, and W3 using OLS, and for each the resulting linear estimation was used to find estimates of the missing 132 133 observations. The estimated equations are presented below. W4 : 12.3 + 0.413 W1 + 0.488 W2 - 0.0812 W3, (1.41) (6.36) (0.943) (-0.603) R-Squared : 0.997, F = 6983. W5 = - 16.9 - 0.0602 W1 + 0.985 W2 + 0.252 W3, (-1.92)(-1.54) (29.6) (2.91) R-Squared = 0.99, F = 29137. W6 2 - 8.53 - 0.0602 W1 + 0.00751 W2 + 1.14 W3, (-1.22) (-1.82) (0.28) (16.6) R-Squared : 0.97, F = 648. W7 : 41.6 + 0.135 W1 + 0.538 W2 - 0.124 W3, (1.20) (0.890) (4.39) (-0.36) R-Squared : 0.98, F = 781. W8 = 47.1 + 0.0947 W1 + 0.531 we - 0.136 W3, (1.29) (0.0593) (4.13) (-0.379) R-Squared : 0.97, F = 629. The measures of factor prices finally settled upon are variants Of W4 (unit labor cost, private sector, non-farm) and W8 (unit non-labor payments, private sector, non-farm). Both series entail unit costs and both refer to the private, non-farm sector. Because the commodities tested are products of the manufacturing sector, a case can be made for using W1 (unit labor cost, manufacturing) rather than W4. There is no comparable series available for non-labor payments, however, and thus W4 was chosen for its symmetry with each varii estir W4 1 Like COUV the 134 with W8. Furthermore, W4 and W1 are highly correlated with each other: variations in W1 "explain" more than 98% Of the variations in W4. Regressing W4 on W1 yields the following estimated equation: W4 = -11.8 + 1.15 W1, (-5.12) (68.2) R-Squared = 0.988, F = 4655. A time series called PL was constructed by converting W4 into real terms through the use of the GNP deflator. Likewise, W8, which is eXpressed in "nominal" terms, was converted into real terms by dividing by the GNP deflator; the resulting variable is given the name PK. APPENDIX 3 Equation No. 1 Node) 8 Dependent RX Constant: 34300 (2.25) UHRUU .0522 (.0265) XRIHF “148 (“3.71) GNP728 '6.41 (-l.52)l PL "3406 (*0.426) PK '5206 (’1014) YHCAP83 YHCAP85 YHCAP86 YHCAP87 Rho = 00586 (5.55) R-Squared 3 03° F: 4958 df(flfll,dflfl) (5,53) 135 Table A-1 Iron oxides and hydroxides, pig-ant grade. 2.19 COIIOdity N00: 2 3 8 D PX 0X -1840 18600 (81.60) (1.24) '00966 -8095 ('00793) ('1042) 3.45 -105 (1018) (“2063) 0.846 (2.66)**i 6.85 15.8 (1.24) (0.189) 00148 ‘5097 (00451) ('1001) 0.324 (1.42) 0.840 0.741 (11.9) (8.48) .15 .20 1.82 2.72 (5,53) (5,53) 2.36 2.33 525 6030 4 D PX '78.? ('00777) .0877 (0.791) .0814 (.0286) 1.94 (0.349) -0.422 (-0.124) .00421 (0.279) 0.886 (14.7) .02 0.218 (5.53) 2.42 5 8 0X 32800 (2.03) 0.774 (0.455) 440 (-3.33) '8043 (-1.74)4 ’2203 ('0.267) '5300 (’1011) 0.435 (1.76)! 0.670 (6.93) .28 3.30 (6,52) 2.21 6 E PX -1960 (‘1.66) '00119 (‘00921) 3.47 (1.18) 0.967 (2067)'C' 6.68 (1.20) 0.454 (0.140) '00123 (”00777) 0.860 (13.0) .15 1.48 (6,52) 2.38 anon 136 Table 8-2 Iron oxides and hymxides, pipent grade. Equation "00 7 Node) C “931:0 Util 0 01.1083 Dependent 0X Constant: -22130 ('00993) m 1.80 (0.509) XRIVF -151 ("048) GNP72‘ ’5053 ("00933) P1. 267 (2 .24) PK 74.6 (1.23) “was MS m m R‘SQHQPEd 3 079 F: 1206 (“UNI-.119“) (5,17) Co-odity a... 525 (.030 8 9 10 C C C .LT.83 £2.83 .GE.83 PX 0X PX 1220 52500 -4350 (1.16) (4.34) (-4.91) '00364 6017 '00535 ('2010) (1038) ('1004) 5.86 -116 -0.648 (3.70) (~2.46) (-0.187) 1004 ‘1707 1054 (3.73)“ (-2.53)*I*(3.00)m -9.57 -120 20.6 (-1.71) (-1.63) (3.82) '8099 '9604 .000 (-3.16) (-1.59) (1.80) .79 .65 .65 12.8 11.7 11.8 (5,17) (5,31) (5,31) 137 Table 8-3 Iron oxides and hydroxides, piglent grade. Equation No. 11 Node) 0 Dependent 0X Constant: 23200 (1.54) VRROU '2023 ('1067) XRIHF '119 ('3006) GNP72$ PL '5.24 (‘00623) PK '0907 ('1000) YHCAP83 YHCAP85 -.0253 (‘0.103) YHCAP86 YHCAP87 “”0 3 00677 (7.07) R-Sguared = .21 P: 2054 d1(flfllyd9fl) (5)53) DU: 2.32 oo-odity No.1 12 13 D E PX 0X 37.9 34500 (.0375) (2.21) .0775 .0195 (0.691) (.00951) -.0935 -149 (‘00326) ('3064) ’6039 (’1049) 1016 ’3506 (0.210) (-0.430) '00210 ’5207 (-.0624) (-1.13) ’00110 '00192 (-0.696) (-.0795) 0.897 0.585 (15.6) (5.54) .02 .30 0.267 3.76 (5,53) (6,52) 2.39 2.19 525 6030 14 E PX -1760 ('1051) ’00944 (-0.773) 3.18 (1.08) 0.873 (2.66)888 6.16 (1.10) 0.193 (.0602) -.0138 (-0.912) .15 1.53 (6,52) 2.37 138 Table A-4 Iron oxides and hydroxides, pigaent grade. Cal-odity No.: 525 6030 Equation No. 15 16 17 18 Model C C C C “93C 0Ut110 01.1085 01.7085 0W0” 0fi085 Dependent 0X PX 0X PX Constant: -32000 1050 67700 -4700 ('1.48) (0.682) (6.08) ('5.95) W 3002 '00525 ’1047 ’00376 (1.05) (-2.57) (-0.358) (-0.129) XRIHF -133 4.63 '177 3.03 (-3.65) (1.79) (-4.07) (0.982) 90728 7079 1045 ‘9080 00933 ('1071)' (4049)” ('1058)” (2014)“ PL 275 -4.83 -101 15.9 (2.37) (-0.588) (-1.56) (3.45) PX 146 -14.0 -194 13.7 (2028) ('3009) (‘3048) (3047) YHCAP83 YHCAP85 YHCAP86 YHCAP87 R'SMl‘ed = 062 070 081 071 F= 7.36 14.8 21.7 12.3 df(nul,den) (5,23) (5,23) (5,25) (5,25) 139 Table 4-5 Printing irks. Ca-odity )b.: 533 2000 Emation Ho. 1 2 3 4 Model B 8 D D Dependent 0X PX 0X PX Constant: -16200 402 -6020 527 (‘3.02) (0.252) ('1.06) (0.385) UAROH ‘1065 ’00479 00667 '00293 (‘1.77) (‘1.52) (1.71) ('2.34) XRIHF 9000 '1600 1301 ’1409 (0.848) ('5.12) (0.894) ('4.31) W728 6.24 0.326 (2.75)!“ (0.478) PL 94.0 16.8 38.5 15.6 (3.07) (1.85) (1.20) (2.09) PK 10.9 3.38 11.1 3.81 (0.860) (0.913) (0.667) (1.01) D781794 '351 “134 24.4 '104 (‘1.37) (‘1.74) (1.28) ('2.50) DUH773 417 '253 912 '232 (1010) ('2015) (206‘) (‘2020) DUH774 '1530 644 '1230 651 ('4041) (600‘) (“3066) (6032) YHCAP83 YHCAP85 YNCGP86 YHCAP87 -.0439 .0185 ('0.332) (0.529) Rho = '0.183 '0.272 0.106 '0.280 (‘1.43) ('2.17) (0.820) (52.24) R‘SQUIPEd 053 056 0‘6 056 3 7014 6077 5036 6060 df(nul,den) (8,50) (8,50) (8,50) (8,50) DU= 1091 1094 2000 1095 5 B GX -16400 ('2.73) ‘1064 ('1066) 9.25 (0.772) 6.24 (2.72)808 94.6 (2.90) 11.1 (0.842) -352 ('1036) 415 (1.09) -1540 ('4035) .00585 (.0495) -0.184 (-1.44) .53 6.23 (9,49) 1.91 6 E PX 17.9 ('00101) ’00433 ("1031) '1501 (-4.29) 0.316 (0.459) 18.4 (1.90) 3.85 (1.01) '134 (’1074) -256 ('2016) 638 (5.93) .0180 (0.510) -0.278 ('2022) .59 7.72 (9,49) 1.94 Equation No. 7 Node) C Capac.Util.: .LT.87 Dependent 0X Constant: -17400 ('1093) UARDH -1.75 (-1.37) XRIMF 9.93 (0.655) CHP72$ 6.21 (2.18)Ii PL 106 (2.14) PX 6.43 (0.332) D781794 -410 (-1.20) DU" 773 465 (1.13) DUH 774 -1590 (”4001) YHCAP83 YHCAP85 YHCAP86 YHCAP87 R-Squared = .52 P: 4076 df(nm,den) (8,35) 140 Table A-6 Printing inks. Con-odity 80.: 533 2000 8 C .LI.87 PX -3370 (’1028) '00530 (-1.43) '1606 (-3.76) 0.716 (0.861) 38.0 (2.64) 9.78 (1.73) -202 ('2004) r332 (“2075) 596 (5.14) .68 9.55 (8,35) 9 10 C C £2.87 £2.87 0X PX -9650 5830 (-1.00) (1.45) -.0507 -0.212 (-.0164) (-0.164) -36.1 5.99 (-0.916) (0.363) 3049 '00723 (0.626) (-0.310) -24.9 -1.08 (-0.484) (-.0501) 124 -33.4 (2.46) (-1.58) .49 .28 1.96 0.783 (5,10) (5,10) 2quation Ho. 1 Model 2 Dependent 0X Constant: 3510 (1.41) UAROU 0.372 (0.806) XRIHF ”1.405 ('20‘6) CNP72$ -0.768 (~0.811) PL -1006 ('00733) PX 0.188 (.0269) D781794 1650 (15.1) YHCAP83 YHCAP85 YHCAP86 YHCAP87 Rho = 00119 (0.920) R-Squared = .95 P: 158 df(nul,den) (6,52) DU= 1093 141 Table A-7 RUbber ce.ent0 Connodity No.: 588 3060 2 8 PX -12600 (’1013) 0.949 (0.744) 26.0 (0.900) 1.50 (0.483) 44.4 (0.791) 41.4 (1.25) -2920 (‘6075) 0.751 (8.73) .58 11.9 (6,52) 2.46 3 D 0X 3820 (1.67) ‘00133 (”00663) '1708 (‘2093) '1000 (-0.765) '2041 ('00367) 1590 (21.4) ’00655 ('1017) 0.0525 (0.404) .95 101 (6,52) 1.94 4 D PX -9280 ("00944) 1.37 (1.52) 18.5 (0.700) 33.4 (0.619) 41.0 (1.24) '2880 (-6.87) ‘00890 (~0.248) 0.759 (8.95) .57 11.7 (6,52) 2.42 5 2 0X 4770 (1.79) 0.211 (0.447) ‘1704 ('2063) -0073? (-0.783) ’1505 (“1003) '1050 ('00221) 1650 (15.4) “00634 (-1.13) 0.0714 (0.550) .95 149 (7,51) 1.93 6 2 PX -12000 ('1005) 0.957 (0.744) 24.7 (0.829) 1.42 (0.448) 42.3 (0.735) 40.8 (1.22) '2920 ('6069) ‘00287 ('1079)' 0.754 (8.82) .58 9.97 (7,51) 2.46 142 table A-8 Mk? cunt 0 Che-odity No.: 503 3060 2quation No. 7 8 9 10 Nude) C C C C th0m1103 01.1087 01.1067 £2.87 002087 Dependent 0X PX 0X PX Constant: 5270 '48000 2290 24400 (1.41) (-4.80) (1.44) (2.74) W014 00660 '0090'4 1026 2046 (0.120) (-0.613) (2.47) (0.860) XRIHF '2201 3606 1104 '4407 (~3.07) (1.92) (1.75) (-1.22) W724 '00830 9054 ‘1040 ’1409 (-0.755) (3.25)!!! (-1.52)* (-2.88)0! PL -14.0 197 -10.4 -25.1 (~0.656) (3.46) (-1.22) (~0.526) PX -1.09 '110 '13.3 -3.82 (-0.121) (4.60) (-1.59) (-.0815) D781794 1670 -3470 (13.0) (~10.1) YHCAP83 YHCAP85 YHCAP86 YHCAP87 R-Squared = .96 .88 .93 .94 F: 144 4600 2605 3404 df(nua,den) (6,37) (6,37) (5,10) (5,10) 143 Table A-9 Neesprint paper. Coenodity 80.: 641 1000 2quation Ho. 1 2 3 4 Node) 8 8 D D Dependent 8X PX 0X PX Constant: 138000 -230 95200 237 (1.03) ('0.422) (0.541) (0.534) UhRUU 79.3 -.0415 1.59 '.0719 (3.53) ('0.728) (0.104) ('1.26) XRIHF 517 '1.58 572 '1.02 (1.46) ('1.15) (1.25) ('0.723) GNP72$ '147 0.200 (‘3.47)!i*(1.30)§ PL '35? 3065 ’42? 2060 (“0.462) (1.48) ('0.430) (1.17) PX '397 '0.722 '551 '1.05 ('0.924) (‘0.481) ('1.01) ('0.718) D781794 YHCAP83 YHCAP85 YHCAP86 YHCAP87 '8.53 .000575 ('2.36)§§*(.0832) Rho = 0.126 0.866 0.400 0.968 (0.975) (13.3) (3.35) (29.7) R’SQUIPEd 3 023 016 016 009 F3 4014 1096 2006 1005 di(nua,den) (5,53) (5,53) (5,53) (5,53) ”'3 1093 2045 2000 2061 5 6 2 2 0X PX 319000 -212 (2029) (”00360) 6105 '00404 (2.79) (-0.702) 128 -1.67 (00356) (-1016) -148 0.198 (-3.78)¢OI(1.27) -1120 3.76 ('1046) (1042) “574 "00701 (-1.42) (-0.463) '9013 '000166 (-2.78)l**(-0.264) 0.106 0.863 (0.820) (13.1) .38 .16 5.39 1.66 (6,52) (6,52) 1.91 2.45 144 Table 4-10 Newsprint paper. Co-odity 80.: 641 1000 Emation lb. 7 8 )bde) 2 E Dependent 0X PX Constant: 111000 -427 (0.791) ('0.797) W 108 °.0830 (3027) (-1042) XRIMF 704 '1.35 (1.79) (“1.01) W720 '191 0.350 (-3.32)m(2.13)1u P1. '347 3.88 ('0.431) (1.54) PX “240 '0.206 ('0.522) ('0.140) D781794 W83 4.42 -.0154 (1.34) (‘2.15)" W85 W86 W87 3 00177 00664 (1.38) (13.2) R‘smm 3 026 023 F3 3034 2054 df(llll,den) (6,52) (6'52) “3 1092 2042 145 time A-11 “print paper. (Io-odity 00.: 641 1000 Bmation 8b. 9 10 11 12 Rode) C C C C Capac.Util.: 01.1083 01.1083 0E0” 093083 Dependent (1X PX 0X PX Constant: 212000 -1460 113000 4380 (00646) (“1049) (00855) ('3076) W 100 0.305 101 -0.129 (1.91) (1.95) (2.06) (-0.950) XRIMF -57.4 -4.28 1670 -4.97 (-0.116) (-2.89) (3.21) (-3.45) W720 -226 -0.198 -130 0.288 02.59)!!! (-0.760) (4.68)“ (1.35)! PL -251 10.2 -1130 12.0 00.143) (1.96) (-1.39) (5.35) PK ‘151 0077 ‘658 2098 (-0.170) (1.79) (-0.988) (1.62) 0781794 W83 W85 W86 W87 R‘smm 3 043 089 .34 000 F: 2058 2609 3018 2402 df(mn,den) (5,17) (5,17) (5,31) (5.31) Sputum » M01 Capac.Ut Menden (instant XRIHF 09721 146 Table A-12 fleusprint paper. Cal-odity No.: 641 1000 Equation No. 13 14 15 16 HOdel C C C C Capac 01.11.1103 01.1007 01.1067 003087 003067 Dependent 0X PX GK PX Constant: 355000 -1870 28000 418 (1.24) (-3.76) (0.136) (0.748) m 6500 00164 '307 .00763 (2.77) (2.71) (-0.541) (.0424) XRIHF 0503 -5057 '1140 '00915 (0.220) (-5.57) (-1.35) (-0.398) W720 '162 ' 00954 6304 '00120 (-3.99)II8(-0.912) (0.531) (-0.370) PL -1170 13.0 -804 0.298 (’1009) (4070) ('00729) (00991) PK ‘55 7002 1460 00562 (-1.39) (5.46) (1.35) (-0.191) D781794 YHCAP83 YHCAP85 YHCAP86 ‘ YHCAP87 R-Smal'ed 3 040 066 027 016 F: 5012 5400 00738 00394 di(nul,den) (5,38) (5,38) (5,10) (5,10) XRIHF W721 1751794 001673 DW702 ”I732 Equation No. 1 Hodel B Dependent (1X Constant: -192000 ("00725) UWRUU 44.4 (1.21) XRIHF 177 (0.251) W725 ‘96 .1 ('1028)** PL 2640 (1.84) PX '816 (“00995) 8781794 DUH673 5830 (0.548) DUH702 43400 (4.08) DUH782 '6310 (“0.590) YHCAP85 3 00516 (4.65) R-Squared .40 = 4021 df(fl“lyd€fl) (8,50) 3 1076 147 Table A-13 Pig iron, including cast iron. Cal-odity No.: 671 2000 2 8 PX .4609 (-0.367) '000191 ('00663) 0.143 (0.402) ‘00120 (0.290) .0777 (0.102) 1.01 (2.35) ’5009 ('6.43) '2079 (-0.354) 100 (12.8) 0.223 (1.76) .82 28.6 (8,50) 1.92 8X -313000 ('1027) 5.86 (0.270) 532 (0.800) 3000 (2.15) -966 (“1022) 4410 (0.402) 46100 (4.20) -7830 ('00692) '1027 (-0.288) 0.403 (4.01) .41 4.27 (0,50) 1.79 4 D PX 21.4 (0.168) -.0186 (~1.62) -0.150 ('0.418) '00121 (~0.169) 0.949 (2.37) 50.6 ('6066) “3025 ('00426) 97.8 (12.9) ‘000498 (-2.05)04 0.200 (1.57) .83 31.1 (8,50) 1.91 5 E 0X '186000 (-0.682) 43.3 (1.12) 157 (0.217) -9502 (-1.23) 2610 (1.77) -816 (-0.984) 5800 (0.540) 43300 (4.04) -6600 (-0.593) '00524 (~0.116) 0.520 (4.07) .40 3.00 (9,49) 1.70 6 E PX 12.1 (.0924) -.0279 (-1.14) '00136 (-0.371) .0184 (0.433) ’00907 (-0.123) 0.908 (2.19) ”5101 (’6063) ’2096 (-0.387) 97.4 (12.6) '000540 (-2.05)II 0.207 (1.02) .03 27.3 (9,49) 1.90 Equation Ho. 7 Node) C capiC0Util03 081085 Dependent 0X Constant: -1180000 (-2.47) UAROU -26.7 (-0.416) XRIMF 820 (0.999) (DP925) 51.9 (0.506) PL 7360 (2.87) PK 1400 (0.989) D781794 DU8673 DUH702 58800 (3.77) DUH782 -6000 ('00386) YHCAP83 YHCAP85 YHCAP86 YHCAP87 R'SQUITEd .70 = 6098 df(nu|,den) (7,21) 148 Table A-14 Pig iron, including cast iron. mm 00.: 071 2000 8 9 10 C C C .LT.85 £8.85 £5.85 PX 0X PX 689 -291000 -193 (2095) ('2024) ('8053) -.00163 63.5 -.0117 (’5020) (1033) ('00252) -.0426 -261 -0.171 (-0.106) (~0.513) (-0.345) -.0435 -78.3 -.00535 (-0.872) (-1.01) (-.0760) -3.58 2202 1.09 (‘20867 (2092) (1049) -0.857 460 1.41 (-1.24) (0.700) (2.21) 3700 -4806 (0.461) (-6.22) '4053 ('00597) 95.9 (12.7) .89 .64 .81 25.4 7.21 16.9 (7,21) (6,24) (6,24) Equation Ho. 11 Model C Capac.Uti).t 0LT087 Dependent 0X Constant: -472000 (’1042) ”ARON 30.2 (0.733) XRIMF 704 (1.02) CflP720 '32.8 (’00465) PL 3940 (2.13) PX '715 ('00814) 0781794 0UH673 3810 (0.224) DUH702 66200 (4.16) DUH782 '13000 ('00834) YHCAP83 YHCAP85 YHCAP86 YHCRP87 R-Squared 3 056 = 5058 d1(nul,den) (8,35) 149 Table A-15 Pig iron, including cast iron. Connodity No.: 671 2000 12 C .LT.87 PX 53.0 (0.275) -.00391 (-0.164) -0.119 (“00297) ‘00251 (-0.616) '00300 ('00281) 0.914 (1.80) -4803 (‘4090) ‘4065 (”00505) 98.4 (10.9) .81 18.7 (8,35) 13 C .GE.87 0X 15600 (0.368) 13.4 (0.984) '134 (-0.773) ‘4300 (1.75)! 424 (1.07) '186 (-O.835) .75 0.10 (5,10) 14 C .CE.87 PX 188 (1.02) .0268 (0.453) 0.159 (0.210) -0.115 (-1.07) '1001 (“1002) 0.789 (0.814) 14.4 (5,10) EmflflflnNO0 1 Hflel B fimflflfln 0X finstnn: 640000 (0.894) “NH” 169 (1.77) XRHF' 802 (0.427) GNP72$ -257 (“1.28)! P1 '696 (0.180) PX '4330 (“1.97) 0781794 DUH682 ‘2980 ('0.109) XWIVINIB XTIJRFXES ‘YHIIAPIMS ‘YHIIRP257 ‘ahCD 3' 05548 (5.03) ‘5'3;QF13HF‘Hd 1: 027 F: 3014 (if (nu-,den) (6,52) “.0: 2006 Concrete reinforcing bars. 150 Table (H6 Co-odity 1b.: 673 2005 2 B PX 194 (0.761) -.00373 (“00114) “1079 (“2068) “00698 (“1028) 2.01 (1.48) “00777 (“1000) 59.1 (6.38) 0.587 (5.57) .51 0.92 (0,52) 1.83 0X 221000 (0.339) 59.4 (1.06) 1610 (0.915) 1130 (0.305) -4740 (“2024) '900 (-.0316) 6.46 (0.608) 0.492 (4.34) .27 3.19 (0,52) 2.13 4 D 7.63 (.0324) “00336 (“1066) “1028 (“2004) 2.67 (2.00) “00880 (“8014) 59.5 (6.38) .00392 (1.05) 0.578 (5.45) .50 0.70 (0,52) 5 E 0X 510000 (0.870) 245 (2.16) 340 (0.214) '446 ('2.12)80 1150 (0.347) -4000 (“2014) “5703 (-.00196) 23.7 (2.02)¢O) 0.354 (2.90) .39 4.73 (7,51) 2.01 6 E PX 143 (0.611) .0285 (0.774) “1080 (“2092) -0.156 (“2009)" 2.51 (1.97) “00626 (“00868) 58.8 (6.34) .00791 (1.96)*! 0.522 (4.70) .55 0.03 (7,51) mr1”) b0 7 Hodel C Capac.Util.: 01.1083 Dependent (1X Constant: ~145000 ('0.120) W '47.2 ('0.244) XRIHF '234 ('0.128) W725 222 (0.690) P]. 4160 (0.642) PK '4520 ( “1037) 0781794 1181682 MN m W86 W87 R'smm = 0“ F: 20“ 151 Table A-17 Concrete reinforcing bars. Cmodity 80.: 673 2005 8 9 10 C C C .LT.83 .CE.83 £8.83 PX (1X PX 496 510000 -136 (1.35) (1.18) (-0.757) 0.190 516 -.0612 (3.24) (3.24) (-0.920) -1.74 243 -1.70 (“3012) (00103) (“2040) -0.475 -948 .0510 (-4.86)N¢(-3.79)!fl(0.487) 1.48 2630 3.44 (0.753) (0.995) (3.11) -1.34 3290 -0.320 (“1034) (“8051) (“00$1) 13600 55.8 (0.447 ) (4.39) .85 .74 .55 19.2 14.0 6.13 (5,17) (0,30) (0,30) Emation No. 1 Node) 8 Dependent (1X Constant: -5220 (-0.869) URRUU 0.204 (0.223) XRIHF -21.5 (“1035) GNP72$ 0.478 (0.266) PL 82.1 (2.42) PK “2604 (“1039) D781794 YHCAP83 YHCAP85 YHCAP86 YHCAP87 Rho 3 00343 (2.81) R-Squared 3 055 P3 1301 «(...-,0...) (5,53) 003 1064 152 Table 0'18 Copper alloy lire, bare. Co-odity 110.: 002 2100 2 B PX -3280 (“1040) “00560 (-1.38) “1600 (“2071) 0.427 (0.564) 36.5 (2.66) 15.4 (2.04) .00728 (.0560) .23 3.13 (5,53) 2.04 8X -1470 (“00240) 0.114 (0.214) “2008 (-1.81) 65.6 (1.89) “2807 (“1051) -0.175 (“1036)3 0.369 (3.05) .55 12.9 (5,53) 1.82 4 D PX -3600 (“1036) “00283 (“1020) “1600 (“2035) 37.8 (2.58) 17.0 (2.25) .0332 (0.502) .0204 (0.157) .22 3.02 (5,53) 2.03 5 E 0X -1840 (-0.280) “000632 (“000677) “2800 (“1065) 0.284 (0.156) 67.0 (1.87) “2901 (“1051) -0.173 ('1.33)! 0.367 (3.03) .55 10.6 (0,52) 1.82 6 E PX -3910 (“1044) “00485 (-1.11) “1504 (-2.23) 0.412 (0.539) 39.1 (2.63) 16.1 (2.07) .0317 (0.476) .0115 (.0883) .23 2.50 (0,52) 2.03 Equation ND. 7 lodel C CapaC.Util.: 011067 Dependent 0X COnStant: -16400 (“2066) UWROU “00469 (“00665) XRIHF ‘27.3 ('2.31) GNP72$ 1.40 (1.13) PL 140 (4.24) PK 3.82 (0.238) 0781794 0UH671 0741742 1150 (5.36) YHCAP83 YHCAP85 YHCAP86 10l333§7 R-Smared 3 063 F3 ”06 df(jlflI,dEfl0 (6,37) 153 Table 0'19 Copper alloy Iire, bare. Cal-odity No.: 8 9 C C .LT.87 .CE.87 PX 0X ~2920 5470 (-0.956) (0.769) -0.595 0.751 (-1.70) (0.342) -12.4 -7.07 (-2.12) (-0.229) 0.623 -0.754 (1.02) (-0.190) 33.1 -32.8 (2.02) (-0.896) 1009 “00554 (1.38) (-.0124) -226 (-0.886) 15.5 (0.147) .22 .65 1.74 2.76 (6,37) (6, 9) 682 2160 10 C .GE.87 PX '4360 (-0.505) 0.936 (0.351) -49.5 (“1032) “2070 ('0.561) 44.8 (1.01) 55.4 (1.02) 517 (1.67) .64 2.72 (6, 9) Equation Ho. 11 Node) 8 Dependent 0X Constant: -10000 (“2012) (MW -.0482 (-.0657) XRIMF -20.9 (-1.70) 99726 1.42 (0.994) P1. 95.6 (3.59) PK “7015 (-0.465) (”1671 -202 (-0.779) 0741794 1140 (5.24) “W83 WBS W86 W87 Rho = 0.278 (2.22) R-Squared = .74 3 2003 0'3 1096 154 Table 4-20 Copper alloy nire, bare. Cal-odity "D0: 662 2160 12 8 PX -2580 (“1020) “00524 (“1045) “1402 (“2054) 0.445 (0.653) 33.2 (2.70) 10.2 (1.43) 665 (4.36) “4023 (“00369) .0432 (0.332) .43 5.44 (7,51) 2.08 13 1) 0X -7500 (“1045) 0.495 (1.13) “2600 (-2.02) 86.0 (3.02) “6047 (-0.405) -186 (-0.694) 1090 (4.90) “00426 (-0.374) 0.299 (2.41) .72 19.0 (7,51) 1.96 14 0 PX -1840 (-0.752) “00342 (“1059) -15.5 (“2053) 30.3 (2.28) 10.5 (1.44) 671 (4.32) “2000 (-0.172) -.0140 (“00232) .0572 (0.440) .42 5.22 (7,51) 2.07 (“1073) -0.101 (-0.133) “2203 (“1066) 1.38 (0.962) 92.6 (3.22) “6010 (“00511) '183 (-0.678) 1120 (5.01) “00342 (-0.301) 0.278 (2.22) .74 17.5 (8 ,50) 1.95 -2280 (“00902) “00552 (-1.42) “1406 (-2.37) 0.445 (0.645) 32.0 (2.37) 9.77 (1.32) 672 (4.29) “6069 (-.0756) -.0136 (-0.fl6) .0458 (0.352) .43 4.67 (8 ,50) 2.09 155 Table 4-21 Copper and copper alloy ponder and flakes. Connodity No.1 682 2400 Eqntion )b. 1 2 Node) 3 B Dependent ax PX Constant: 5920 1110 (1.42) (0.592) VGRUU '0.292 '0.319 ('0.521) ('1.16) XRIHF 13.2 '4.73 (1.20) ('0.955) GNP720 1028 “00425 (1.09) (-.0775) PL '26.1 13.8 ('1.16) (1.31) PX '34.8 '8.72 ('2.71) ('1.48) 0781794 DUH681 '310 141 ('1.92) (1.64) YHCRPO3 YHCAPBS YHCAP86 YHCAP87 3 00537 00414 (4.89) (3.49) R'Stlfll‘ed = .27 .19 F: 3.27 2.04 DU= 1.82 2.17 0X 7290 (2.03) 0.356 (1.14) 13.0 (1.36) “2903 (-1.42) “3709 (“3028) -321 (“1095) 0.117 (1.82). 0.440 (3.76) .37 5.01 (6,52) 1.86 4 D PX 1550 (0.846) “00364 (“2040) “S077 (“1017) 11.4 (1.08) “0066 (-1.47) 143 (1.69) “40303 (“00920) 0.435 (3.71) .20 2.12 (4,52) 2.21 5 E 0X 4150 (0.978) -0.152 (“00272) 18.2 (1.62) 1.37 (1.18) “1801 (-0.797) “3502 (~2.76) -318 (“2002) 0.112 (1.74)! 0.554 (5.11) .31 3.21 (7,51) 1.88 6 E PX 1550 (0.791) “00385 (“1034) “5076 (“1010) .000739 (.00132) 11.4 (1.06) “0060 (-1.44) 143 (1.68) “00303 (-0.908) 0.435 (3.71) .20 1.73 (7,51) 2.21 Equation No. 7 Hodel C CipBC4Ut110: 0LT066 Dependent 0X Constant: 17100 (3.57) VHROU 0.643 (0.969) XRIHF 12.6 (1.29) GNP72‘ “1003 (“00924) PL '83.7 (“3015) PK “5505 (“4023) 0781794 DUH681 YHCAP83 YHCAP85 YHCAP86 10(3'457 R'squ3PEd = 042 F: 4031 df(llll,dflfl) (5,30) 156 Table 4'22 Copper and copper alloy ponders and flakes. Coneodity Ho.: 682 2400 8 C .LT.86 PX -2830 (“1016) “00216 (“00637) “4060 (-0.920) .0191 (.0336) 38.4 (2.83) “3085 (-0.576) .42 4.29 (5,30) 9 C .CC.86 0X 3820 (1.33) 0.863 (0.668) 17.9 (1.46) “00331 (“00166) 0.334 (.0188) “4500 (“3010) -301 (“1092) .86 17.5 (6,17) 10 C .GE.86 PX 3810 (2.46) '1.70 (“2043) “1006 (“1060) 1.98 (1.83)** “3047 (-0.361) “1307 (“1074) 175 (2.07) .47 2.55 (6,17) Aluainul and aluainua alloy lire, not insulated. Equation "00 1 Hodel B Dependent 0X Constant: -77700 (“1088) UhROU 14.4 (1.99) XRIHF 167 (1.55) CNP725 '5.91 (“00443) PL 296 (1.24) PK 142 (1.07) 0781794 YHCAP83 YHCRPBS YHCAP86 YHCAP87 R-Squared = 029 F: 4044 df(nul,den) (5,54) 0": 1090 Coo-odity Ho.: 684 2140 2 B PX -512 (“00360) -0.282 (-1.13) “6077 (“1082) 0.215 (0.467) 12.2 (1.49) 2.56 (0.561) .10 1.14 (5,54) 1.86 157 Table A-23 3 D GX -76400 (“1079) 10.6 (2.68) 150 (1.23) 300 (1.27) 125 (0.960) “00359 (“00420) .29 4.43 (5,54) 1.92 4 D PX -380 (“00259) “00169 (-1.23) “6083 (“1062) 11.5 (1.41) 3.05 (0.678) .00375 (0.127) .09 1.10 (5,54) 1.86 5 B OX -74800 (“1072) 13.0 (1.38) 153 (1.24) “4023 (-0.278) 289 (1.20) 136 (0.992) “00234 (-0.241) .29 3.64 (6,53) 1.91 6 E PX -472 (-0.316) “00302 (“00935) “6097 (“1064) 0.238 (0.455) 12.1 (1.46) 2.46 (0.524) “000328 (“000328) .10 0.938 (6,53) 1.86 158 Table A-24 Aluninuo and alminul alloy sire, not insulated. Bmation lb. 7 Hodel C WEAR“ 0: 01.1065 Dependent 0X Constant: -224000 (-1.60) W 17.3 (0.928) XRIlE' 199 (0.844) W723 -2.35 (“00799) P1. 1110 (1.448) PX 418 (1.01) D781794 W83 W85 W86 W87 R-Smared = .23 F3 1036 df (numden) (5,23) Co-odity Mm: 684 2140 8 9 10 C C C .LT.85 £8.85 £8.85 PX (1X PX 3430 -16400 -601 (1.19) (-0.816) (-0.28'5) 00906 5012 “00409 (0.236) (0.694) (-0.527) -1.07 103 -12.0 (“002220) (1032) (“1046) “00235 2048 00203 (-0.388) (0.222) (0.172) ~10.6 -1.17 16.3 (-0.690) (- .0150) (1.33) “1006 6037 4032 (-1.27) (.0834) (0.409) .12 .51 .14 0.601 5.14 0.804 (5,23) (5,25) (5,25) 159 Table A-25 Alulinua and alulinul alloy ponder and (lakes. Equation "00 1 node) 8 Dependent 0X Constant: 2420 (0.127) URRUU “1090 (“00664) XRIHF '53.8 (“1007) CNP72$ 8.26 (1.49) PL “1066 (“00157) P‘ “2202 ('0.374) 0781794 DUH762 3840 (4.48) YHCAP83 YHCAP85 YHCAP86 YHCRP87 Rho 3 00439 (3.76) R-Squared = .51 = 6094 01mm...) (6,52) DU= 1079 Cal-odity “0.: 684 2420 2 8 PX -839 (“00260) 0.253 (0.579) “00931 (-0.118) “00642 (-0.968) 2.16 (0.130) 16.6 (1.78) -203 (“1051) 0.443 (3.79) .25 2.09 (6,52) 2.02 3 D 0X 3100 (0.178) 2.83 (1.83) “4800 (-1.01) “1004 (-.0104) “6088 00.123) 3700 (4.25) 0.527 (1.62) 0.381 (3.16) .54 10.1 (6,52) 1.92 4 D PX '2770 (-1.02) “00137 (-0.566) “00375 (“00506) 12.5 (0.792) 17.4 (2.00) -172 (~1.24) .0150 (2.93)‘** 0.364 (3.00) .29 3.60 (6,52) 1.99 5 E 8X -3570 (-0.191) “00640 (“00225) “3305 (-0.670) 7.41 (1.37) 23.3 (0.224) “2107 (-0.378) 3840 (4.46) 0.460 (1.42) 0.410 (3.45) .54 0.55 (7,51) 1.87 6 E PX -1110 (“00362) 0.299 (0.657) -0.127 (“00155) “00669 (-0.988) 3.48 (0.204) 16.5 (1.76) (-1.50) .0219 (0.422) 0.441 (3.78) .25 2.47 (7,51) 2.00 160 Table 4-26 Alminul and alulinul alloy ponder and flakes. Mity n... 684 2420 Emation No. 7 8 9 10 llodel C C C C “WCOUtiIG: 01.1086 01.1066 005066 093086 Dependent ox 9x ax 9x Constant: -28800 -2290 -9490 -2000 (-1.12) (-0.964) (-1.05) (~0.603) W 2.72 0.356 -2.78 -1.52 (0.747) (1.06) (-0.683) (-1.02) XRIHF 44.0 -5.98 -170 -9.52 (0.836) (4.23) (-4.44) (-0.673) W728 5.27 -O.733 9.96 1.37 (0.876) (4.32) (1.58) (0.590) P1. 104 16.4 62.6 19.9 (0.722) (1.23) (1.12) (0.966) P‘K 28.7 14.5 95.2 11.5 (0.411) (2.25) (2.12) (0.696) 0781794 “.8962 3990 '140 (3.34) ( '1.27) W83 “W85 W86 WU R-Smared ’- 056 023 094 069 -'- 6028 1.45 5603 8.00 “(WARD (6,29) (6,29) (5,18) (5,18) 161 Table A-27 Zinc and zinc alloy sheets, plates, and strip. Miw “30: 686 3220 Emation lb. 1 2 3 4 5 6 lbdel B B D I) E E Dependent 0X P‘X (1X PX 0X PX Constant: 6690 -884 6720 -708 9350 414 (0.585) (-0.970) (0.658) (-0.771) (0.770) (-0.439) 0112011 0.215 .0636 0.266 -.o151 0.541 .0705 (0.156) (0.591) (0.275) (0.166) (0.411) (0.497) 121112 26.6 0.27 33.7 -3.51 27.1 -4.06 (0.903) (-1.36) (1.25) (-1.47) (0.665) (-1.66) | W73 “00629 “00154 “1024 ~0.200 ' (0.269) (0.560) (0.364) (0.759) 1 P1. “3905 9046 “4602 0046 “5506 705 (0.674) (1.63) (0.657) (1.63) (0.932) (1.40) PK “1706 40v “2207 3066 “2204 3092 i (0.527) (1.46) (0.672) (1.31) (0.659) (1.42) 1 0761794 111311263 ‘ 1 11162265 1 11121266 111311267 0.326 -.0175 0.326 -.0209 I 02.01)» (0.657) (0.96111 (-1.02) ' 16.6 = 0.666 0.310 0.791 0.261 0.760 0.242 (7.27) (2.50) (9.94) (2.25) (9.58) (1.92) R-Sqw‘ed = .05 .12 .12 .13 .12 .14 3 00534 1039 1041 1054 1014 1046 df(nun,den) (5,53) (5,53) (5,53) (5,53) (6,52) (6,52) W: 2.40 2.02 2.42 2.02 2.40 2.02 162 Table A-28 Zinc and zinc alloy sheets, plates, and strip. liq;na1:icnn '40. 7 Mel C Capac .Utl) 0: 01.1067 Thepenndenvt 0X Constant: 12700 (1.25) W -1.88 (-l.53) XRIHF 41.0 (-0.544) 019728 2.36 (1.11) PL “6506 (-1.16) ("X -19.7 (“00752) 0781794 W83 W85 W86 W87 R-Sq1ared = .10 F: 00636 (11.10,...) (5,361 (to-66119 116.: 666 3220 8 9 10 C C C .LT.87 .C£.87 .GE.87 PX 0X PX 338 31700 -1160 (0.288) (2.97) (-1.33) 00694 “00705 “00149 (00468) (“00206) (“00535) -6.72 ~97.3 3.68 (“2066) (“2022) (103) -0.379 -9.53 0.692 (“1054) (“1054). (1036)“ 5.70 -79.6 0.566 3.09 2.67 4.91 (1.02) (0.476) (1.08) .24 .86 .62 2.38 12.4 3.15 (5,36) (5,10) (5,10) 163 Table 0-29 Door and 01111100 sash, frales, louldinq, and trio of iron and steel. Co-odity No.: 691 1020 Marion lb. 1 2 3 4 5 6 Node) 8 8 D l) E E Dependent 0X PX 0X PX 0X PX Constant: '15600 '1680 -14600 -2610 -1350 0030 (-0.707) (-0.725) (-0.750) (-1.41) (-.0681) (-1.40) W01) 10.5 -0.115 7.15 -.0783 10.0 -0.141 (3.69) (-0.450) (4.20) (-0.460) (3.72) (-0.584) XRIHF “3102 “2060 “4101 0W2 “6901 00686 (-0.540) (-0.471) (-0.804) (.00616) (~1.31) (0.158) W73 “5087 “00244 “7050 00203 (-0.964) (-.0392) (-1.36) (0.355) P1. -10.7 21.9 -32.2 26.7 -78.1 27.9 (-.0912) (1.92) (-0.293) (2.60) (-0.728) (2.58) PX 141 -0.805 146 -1.17 148 -0.895 (2.11) (-1.22) (2.30) (-0.191) (2.46) (-0.145l 0781794 W83 W85 “(29286 -0.872 .0772 -0.914 .0787 (-2.81)l*§(2.84)m (-2.94)&H(2.85)0H M2287 “)0 3 00566 00760 0.608 00757 00526 00744 (5.56) (9.57) (5.88) (8.90) (4.76) (8.54) R'm ‘ 049 010 053 .22 062 022 7: 1004 1014 1200 2091 1402 2041 “(ll-M) (5,53) (5 ,53) (5,53) (5,53) (6,52) (6,52) W 2.00 2.52 2.18 2.58 2.08 2.58 164 Table A-30 Door and 01118011 sash. frales, ooulding, and tri- of iron and steel. mm 116.: 691 1020 Equation 11... 7 6 9 10 Hodel C C C C 0393(0).”110: 01.7066 01.1006 003066 003066 Dependent 0X PX 0X PX Constant: 37 00 “4660 31300 34820 (0.148) ('2.84) (3.51) (’2.20) W 16.2 '0.866 16.6 '1.69 (4.66) ('3.82) (4.13) (“1.72) XRIHF '51.9 '0.938 36.8 -23.4 ('1.01) (“0.280) (0.971) ('2.52) W725 '15.5 1.70 '23.7 2.51 (“2.67)m(4.49)m ('3.80)l§§(1.64)i P1. '134 31.6 '161 34.9 (“0.961) (3.48) (32.92) (25.7) PK 160 3074 “4504 2006 (2033) (00837) (“1002) (1090) 0781794 W83 W85 W86 “W87 R-Smred 3 077 049 092 050 P3 1909 5077 3807 3054 41(ml,den) (5,30) (5,30) (5,18) (5,18) 165 Table A-31 Door and linden sash, frames, Ioulding, and trin of aluninun. Equation "00 1 Node) 8 Dependent 0X Constant: “16700 (“1.05) UfiROU 1.20 (0.596) XRIHF “12.4 (“0.299) GIF72$ 6.56 (1.51) P1 73.2 (0.875) PK 20.6 (0.431) 0781794 YHCAP83 YHCAP85 YHCAP86 YHCAP87 Rho 3 00606 (5.85) R-Squared 3 043 F3 7097 d14fl“l,den) (5,53) 1.86 mm a... 691 2020 2 8 PX “2210 (“0.692) .0829 (0.250) 3.44 (0.430) 0.940 (1.03) 10.8 (0.712) 3.28 (0.376) 0.875 (13.9) .04 0.442 (5,53) 1.73 3 D 0X 6540 (0.429) 2.62 (1.87) '4807 ('8021) '5031 (“.0628) '1302 (“0.261) '00436 ('8094)" 0.750 (8.72) .28 4.09 (5,53) 4 0 PX “508 (“0.190) 0.259 (0.941) 0.690 (.0925) 6.65 (0.456) 2.91 (0.330) .0138 (0.361) 0.863 (13.1) .02 0.253 (5,53) 1.71 5 8 0X 5880 ('00347) 1.23 (0.617) -2300 ('00629) 4.88 (1.08) 31.8 (0.366) 3.53 (.0717) '00391 ('8072)*' 0.685 (7.22) .36 4.95 (6,52) 1.92 6 E PX “2420 (“0.746) .0665 (0.199) 4.44 (0.534) 0.944 (1.04) 12.0 (0.774) 3.13 (0.355) .0160 (0.418) 0.867 (13.4) .04 0.390 (6,52) 1.70 166 Table MR Door and linde- sash, frales, Iouldinq, and tri- of aluninun. Co-odity 1b.: 691 2020 Emation lb. 7 8 9 10 lbdel C C C C WOUtil0: 01.1086 01.1056 04:30“ 0W0“ Dependent 0X PX 0X PX Constant: “40400 “7430 “12900 -527 (“2.13) (“2.16) (“1.84) (“0.308) W '00482 ’00698 '4010 '1017 (“0.183) (“1.47) (“1.30) (“1.53) ”1110' “21.6 4.26 “79.0 “15.3 (“0.558) (0.608) (“2.65) (“2.11) 98724 13.4 1.93 12.2 1.25 (3.05)“ (2.42)“ (2.50)“ (1.05) Pl. 161 47.2 38.7 12.2 (1.53) (2.48) (0.894) (1.15) P! 84.6 6.32 77.8 12.8 (1.63) (0.674) (2.23) (1.50) D781794 W83 W85 W86 W87 R“Smared = 079 029 089 005 P: 2308 2050 2809 6061 Equation No. 1 Node) 8 Dependent 0X Constant: 3090 (0.200) UhROU “2.10 (“0.855) XRIHF “71.5 (“8075) CNP72$ 1.53 (0.324) PL 116 (1.31) P‘ “5500 (“1.12) 0781794 YHCAP83 YHCAP85 YHCAP86 YHCAP87 = 00259 (2.06) R-Squared = .24 3 3036 81(nul,den) (5)53) DU= 8096 167 Table A-33 Cal-odity "0.: 695 3140 PX “1750 (“2056) .0261 (0.239) “00990 (“00545) 0.105 (0.502) 8.06 (2.05) 7.69 (3.52) 0.262 (2.08) .21 2.81 (5,53) 2.06 “3880 (“00257) “00338 (“0.250) “4508 (“8080) 138 (1.61) “4607 (“0.994) 0.482 (1.71). 0.253 (2.01) .28 4.17 (5,30) 1.92 4 D PX “1410 (“8099) .0505 (0.801) “8062 (“00934) 6.75 (1.68) 7.64 (3.46) “00806 (“0.812) 0.294 (2.36) .20 2.65 (5,53) 2.07 Hacksan blades, hand and paper. 5 B OX “3280 (“00208) 0.214 (.0768) “4609 (“8009) “8009 (“00222) 135 (1.54) “4400 (“00897) 0.503 (1.67)! 0.263 (2.09) .28 3.34 (6,52) 1.92 6 E PX 1560 (“2088) “00389 (“00254) “8060 (“0.821) 0.170 (0.762) 7.38 (1.84) 7.36 (3.30) “00833 (“0.976) 0.276 (2.21) .22 2.40 (6,52) 2.07 168 Table A-34 )lacksau blades, hand and poser. Co-odity 1b.: 695 3140 Equation 80. 7 8 9 10 Node) C C C C “DEC 00(8) 0 8 08.8085 08.1065 003085 008005 Dependent 0X PX (IX P'X Constant: “109 “3170 1990 “1300 (-.00331)(-2.61) (0.138) (-1.72) m 8088 “00460 “3096 .0295 (0.269) (“0.285) (“0.750) (0.106) XRIHF “38.6 “0.149 “36.3 “5.07 (“0.697) (“.0729) (“0.647) (“1.72) ”72‘ “2085 00356 6092 “00845 (“0.413) (1.40) (1.11) (“0.343) P1. 160 14.9 41.1 8.61 (0.910) (2.30) (0.492) (1.95) PK “9905 9098 “4300 0077 (“1.02) (2.78) (“0.598) (2.32) D781794 W83 W85 W86 “W87 R“Smared = 036 040 058 042 F: 2.60 3080 S086 3064 Emation Ho. 1 Node) 8 Dependent 0X Constant: -6610 (“8065) UARUU .0833 (0.181) XRIHF “2.22 (“0.214) GNP720 3.02 (2.81)!!! P1. 28.6 (1.42) PX 0.409 (.0351) D781794 YHCAP83 YHCAP85 YHCAP86 YHCAP87 Rho 3 00727 (8.14) R“Squared = 050 3 8004 «(...-,0...) (5,53) 2.18 101st drills, natal-cutting. 169 Table A-35 Co-odity a... 695 4145 2 PX 29800 (2.30) “2002 (“8007) “8404 (“00423) “4.67 (“8024) “114 (“8056) “2709 (“00667) 0.408 (3.43) .53 12.1 (5,53) 1.90 3 D 0X 1180 (0.316) 0.819 (2.13) “8609 (1.64) 1.68 (.0822) “2098 (“00235) “00796 (“8037), 0.858 (12.8) .21 2.78 (5,53) 2.25 4 D PX 17900 (1.39) “3034 (“3008) 13.9 (0.417) “6604 (“0.938) 30.3 (“0.756) 0.390 (1.49)! 0.409 (3.44) .54 12.4 (5,53) 1.89 5 E 0X “5380 (“8030) 0.119 (0.260) “4066 (“0.436) 2.81 (2.55)OI 23.0 (1.12) “0.198 (“00869) “00628 (“8007) 0.751 (8.75) .47 7.64 (0,52) 2.25 6 E PX 20800 (1.59) “8044 (“00766) “4063 (“0.137) “4063 (“1.28) “7003 (“00963) “8407 (“0.381) 0.399 (1.52)! 0.354 (2.90) .60 12.8 (6,52) 1.88 170 Table A-36 Twist drills, Ietal-cutting. oa-odity Mm: 695 4145 Equation lb. 7 8 9 Hodel C C C Capac.Util.: .LT.87 .LT.87 £3.87 Dependent 0X PX (D( Constant: “8510 -9780 559 (“2.33) (“0.814) (0.236) UARW -0.887 1.69 1.04 (“2.00) (1.16) (1.38) XRIMF “13.1 23.6 2.61 (-1.79) (0.980) (0.269) W720 4.43 “6.47 “1.19 (5.79)m (“2.56”) (“0.868) P1. 47.4 64.9 0.195 (2.34) (0.972) (.0154) P! 8060 5903 “4046 (0.170) (1.91) (“0.360) D781794 W83 “W85 W86 W87 R-Saned = .90 .72 .82 P= 70.9 19.4 9.18 df(m|,den) (5,38) (5,38) (5,10) 10 C $8.87 PX 82200 (2.97) 1.80 (0.202) “144 (“8026) “3007 (“8098). “118 (“0.798) “159 (“8009) .84 11.0 (5,10) 171 Table A-37 Safety-razor blades. Co-odity Mr... 696 0340 Emation lb. 1 2 3 4 Rude) 8 8 D D Dependent 0X PX (1X PX Constant: “410000 “33700 “234000 “160000 (“1.11) (“0.202) (“0.544) (“1.16) UARUU “808 “8802 2202 “2002 (“1.60) (“0.558) (0.559) (“1.64) XRIHF “1410 “802 “1310 “50 (“1.45) (“1.85) (“1.10) (“0.964) W720 243 “43.8 (1 .94)’ (“0.962) PL 3840 1170 2900 1700 (1.77) (1.38) (1.16) (2.18) PX “322 431 46.5 238 (“0.273) (0.866) (.0356) (0.513) 0781794 W91 271000 “22600 287000 “27900 (8.20) (“4.23) (7.91) (“5.20) “W83 W85 “W86 3.32 5.93 (0.314) (2.64)!" W87 Rho 3 “00893 00702 “000668 00664 (“1.51) (7.56) (“0.508) (6.81) R'squ1PEd 074 036 070 042 3 2403 4064 2007 6033 df(lll,d€fl) (6,52) (6,52) (6,52) (6,52) DH= 8060 2082 8055 2080 5 E 0X “374000 (“00960) “123 (“8045) “1540 (“8044) 260 (1.88)! 3690 (1.64) “348 (“0.291) 275000 (7.66) “3028 (“00297) “0.192 (“1.51) .74 20.5 (7,51) 1.60 6 E PX “115000 (“0.714) “8206 (“0.658) “518 (“8023) “2506 (“0.591) 1570 (1.91) 284 (0.598) “27300 (“5008) 5.75 (2.52)!“ 0.692 (7.35) .43 5.46 (7 ,51) 2.14 Equation Ho. 7 Hodel C CQDBC0Ut810: 088066 Dependent 0X Constant: 153000 (0.198) ”ARCH 11.3 (0.106) XRIHF “513 (“0.326) CNP72$ 119 (0.664) PL 844 (0.197) PK “2790 (“8032) D781794 DUH791 YHCAP83 YHCAP85 YHCAP86 YHBRP87 R“Squared 3 045 F: 4095 «(II-fie“) (5,30) 172 Table A“38 Safety-razor blades. Chi-odity “D0: 8 9 C C 081066 006066 PX GX “41300 306000 (“0.286) (0.949) “4302 “274 (“2.17) (“1.93) “55.4 “982 (“00866) (“00762) 27.0 454 (0.810) (1.99)uu 1150 “2050 (1.44) (“0.995) “405 “759 (“1.03) (“0.512) 330000 (15.8) .50 .96 5.91 78.7 (5,301 (6,17) 696 0340 10 C .CE.86 PX -390000 (“3039) 31.6 (0.622) 1700 (3.71) 64.0 (0.787) 521 (0.710) 738 (1.40) “29800 (“3099) .77 9.41 (6,17) Emation lb. (lode) Dependent Constant: W XRIMF W724 P1. PX D781794 W83 W85 W86 W87 Rho = R“Sq1ared 08(HUI’dED) 173 Table 4-39 l'lotors, AC, polyphase--induction, not over 20 hp. Co-oditv lb.: 716 4042 1 2 3 4 5 B B D D E 0X PX (IX PX (1X “78600 “479 “15000 “732 “39100 (“0.751) (“0.990) (“0.166) (“1.52) (“0.389) 8403 “00833 8206 “00838 7023 (1.10) (“1.64) (1.58) (“2.97) (0.575) “270 “0.776 “428 “.0526 “366 (“0.986) (“0.606) (“1.82) (“.0384) (“1.40) 8.76 “.0467 15.2 (0.306) (“0.305) (0.558) 332 5.52 65.1 6.41 133 (0.605) (1.97) (0.128) (2.35) (0.254) 372 1.84 369 1.93 362 (1.18) (1.19) (1.25) (1.31) (1.22) “404 00804 “4088 (“2078)"*(8080) 0.629 0.119 0.642 0.0962 0.636 (6.22) (0.920) (6.43) (0.742) (6.34) .27 .66 .34 .68 .35 3.85 20.3 5.51 22.2 4.70 (5,53) (5,53) (5,53) (5,53) (6,52) 2.08 2.00 2.04 2.00 2.02 6 8 PX “657 (“8083) “00579 (“00586) “.0171 (“00823) “00835 (“0.816) 5.99 (2.14) 2.22 (1.41) .0138 (“2.73)m(1.33)l 0.118 (0.911) .67 17.5 (5,52) 1.98 174 Table A“40 Motors, MI, polyphaseuinduction, not over 20 hp. Equation "00 7 lode) C Capac.lltil .: .LT.85 Dependent 0X Constant: “162000 (“00728) W 8.04 (0.270) XRIMF “425 (“1082) W728 31.8 (0.675) P1. 654 (0.545) PR 710 (1.07) D781794 W83 W85 W86 W87 R-Smared = .58 F: 6042 df(m,den) (5,23) Co-odity 1b.: 716 4042 8 9 10 C C C .LT.85 .GE.85 £3.85 PX (IX PX “1150 1360 “399 (“1.40) (.0329) (“0.635) “00210 7027 “00358 (“00891) (00470) (“00855) 1.80 “185 “3.44 (1.29) (“1.14) (“1.40) “.0766 8.84 “0.320 (“00442) (00383) (“00913) 7079 “9020 6050 (1.76) (“.0381) (1.78) 2.02 55.2 3.81 (0.826) (0.266) (1.21) .65 .84 .74 0.40 26.4 14.2 (5,23) (5,25) (5,25) Equation llo. node) Dependent Constant: XRIMF GNP72$ PL PX D781794 YHCAP83 YHCAP85 YHCAP86 YHCAP87 R-Squared = P: df(nul,den) 175 Table A“41 llotors, AC, polyphase-“induction, not over 20 hp. E9 #- H 0X “43500 (“00456) 15.4 (1.85) “356 (“1.44) 192 (0.359) 370 (1.19) “1070 (“1.03) 0.629 (6.22) 4.11 (5,53) 2.10 Co-odity lb.: 716 4042 12 D PX “1300 (“2083) “00643 (“2006) 0.696 (0.587) 9.20 (3.59) 2.60 (1.96) .0367 (3.21)!!! .0460 (0.354) .74 30.9 (5,53) 2.00 13 B GX “51600 (“0.477) 13.8 (1.05) “335 (“1.19) 4.79 (0.166) 217 (0.387) 367 (1.17) “8067 (“0.985) 0.629 (6.22) .28 3.37 (6,52) 2.09 14 E PX “1260 (“2064) “00665 1'00667) 0.649 (0.531) “00367 (“0.271) 9.02 (3.42) 2.67 (1.94) .0368 (3.18)I0§ 0.742 (0.406) .74 25.0 (6,52) 2.00 Equation No. Hode) Capac.Util .: Dependent Constant: UhRUU XRIHF PL PX D781794 YHCAP83 YHCAP85 YHCAP86 YHCAP87 R“Squared = f: «(...-.02..) 176 Table A-42 Hotors, AC, polyphase-“induction, not over 20 hp. Cal-odity Ho.: 716 4042 15 16 17 18 C C C C .LT.87 .LT.87 .GE.87 .GE.87 0X PX 0X PX “133000 “991 73400 “1010 (“101”) (“8063) (1032) (“00773) N01 “00645 N00 “0W11 (1.87) (“0.875) (1.68) (“.0218) “196 0.704 “125 1.88 (“0.737) (0.578) (“0.546) (0.351) “5083 “00346 “3703 “00205 (“0.209) (“0.272) (“1.16) (“0.271) 409 7036 “443 9071 (0.554) (2.18) (“1.48) (1.39) 666 2.06 57.5 0.129 (1.94) (1.31) (0.197) (.0188) .57 .62 .90 .77 9.92 12.5 19.2 6.58 (5,38) (5,38) (5,10) (5,10) 177 Table A-43 Conbines, self-propelled. Connodity Ho.: 721 2220 Equation No. 1 2 3 4 5 Node) 8 8 D D E Dependent 0X PX 0X PX 0X Constant: “16500 “65900 “14400 “69800 “16100 (“2.14) (“3.19) (“1.89) (“3.34) (2.07) URRDU “2088 “6023 “00366 6082 “3057 (“2.29) (“1.73) (“0.545) (3.29) (“2.01) XRIMF “61.9 “139 “64.8 “66.5 “65.8 (“3.04) (“2.56) (“2.98) (“1.10) (“3.03) GNP724 4.90 19.3 6.00 (2.05)'! (2.89)*** (1.92)! PL 94.0 408 85.4 378 94.2 (2.12) (3.38) (1.95) (3.15) (2.11) PK 80.3 187 93.4 240 77.0 (3.26) (2.80) (3.86) (3.65) (3.04) D781794 YHCAP83 0.123 1.39 “.0983 (00896) (3058)'** (“00558) YHCAP85 YHCAP86 YHCAP87 “IO 3 00206 “00350 00872 .0354 00207 (1.62) (-o.2(9) (1.34) (0.272) (1.63) R-Squared = .34 .57 .31 .50 .34 r: 5.36 14.2 4.77 14.6 4.05 df(nul,den) (5,53) (5,53) (5,53) (5,53) (6,52) DU= 8062 8096 8085 8097 8088 6 E PX “70400 (“3034) 5.15 (0.895) “6902 (“1.13) 1.70 (0.175) 383 (3.14) 236 (2.35) 1.32 (2.35)!** .0261 (0.201) .58 12.1 (6,52) 1.98 Equation Vb. 7 Model C C3pac.Util.: 0LT083 Dapendent 0X Constant: “10700 (“00667) UAROU “2.99 (“1017) XRIMF “56.9 (“2.34) CNP72$ 6.07 (1.42) PL 40.9 (0.476) PK 74.6 (1.71) 0781794 YHCAP83 YHCAP85 YHCAP86 YHCAP87 R-Squared = .58 P: 4062 dflmmden) 15,171 178 Table A-44 Conbines, self-propelled. 00.0111“) 110.: 721 2220 8 9 10 C C C 0LT083 0GE083 063063 PX 0X PX “72200 “18100 “66500 (“1005) (“2035) (“3065) 1004 “4054 7087 (0.949) (“1.61) (1.17) “5504 “8306 “7068 (“0.534) (“2.78) (“0.100) “2.72 5.88 2.18 (“0.149) (1.32) (0.207) 522 150 246 (1.42) (3.21) (2.22) 66.4 57.9 283 (0.356) (1.50) (3.09) .62 .39 .70 5.46 3.92 14.1 (5,17) (5,31) (5,31) 179 Table A-45 Colbines, self-propelled. Co-odity Hm: 721 2220 Equation Ho. 11 12 13 14 Node] C C C C “M001.” 0: 0LT083 013085 03.3086 01.1067 Dependent PX PX PX PX Constant: “72200 “57000 “48600 “65400 (“1.05) (“1.12) (“1.35) (“1.79) m 100‘ 2093 “00766 “5.00 (0.949) (0.433) (“0.158) (“1.13) XRIMF “55.4 “90.2 “102 “146 (“00534) (“1005) (“10%) (“20W) GNP72$ “2.72 7.58 11.8 17.7 (“0.149) (0.709) (1.41)& (2.31)*§ PL 522 446 352 433 (1.42) (1.64) (1.76) (2.14) PK 66.4 34.5 78.2 162 (0.356) (0.230) (0.796) (1.72) D781794 YHCAP83 YHCAP85 YHCAP86 YBCAP87 R“Smared 3 062 069 062 060 P: 5.46 10.3 9.73 11.4 180 Table A-46 Dozers, for lountinq on tractors. mm lb.: 723 4052 Emation lb. 1 2 3 4 s a Hodel a a n o 13 a Dependent ax PX ax PX ax PX Constant: 2700 12.2 “771 1560 4900 “988 (0.444) (.000663) (“0.105) (.0954) (0.702) (“.0503) UAROU 1.43 “0.377 “1.51 2.14 1.24 “0.867 XRIMF “25.4 2.92 “18.6 “10.5 “30.7 10.8 (“1.58) (.0608) (“0.985) (“0.248) (“1.12) (0.213) W724 “5.88 4.92 “5.95 5.83 (“3.00)*!i(0.986) (-3.03)uaa(1,12) PL 54.2 “28.8 65.9 “38.0 44.7 “17.5 (1.54) (“0.302) (1.60) (“0.413) (1.16) (“0.177) PX “5.98 18.8 “19.0 46.3 “7.14 4.89 (“0.306) (0.343) (“0.878) (0.876) (“0.353) (.0870) D781794 YHCAP83 YHCAP85 YHCOP86 W87 “00317 00994 “00116 00204 (“0.672) (0.342) (“0.718) (0.713) Rho = 0.118 0.663 0.158 0.594 0.142 0.705 (0.912) (6.80) (1.23) (5.67) (1.10) (7.64) 10ml!!! 3 034 007 021 010 034 0“ 3 5046 00617 2067 8020 4041 00562 «(n-,den) 15,53) (5,53) (5,53) (5,53) (6,52) (5,52) "3 10“ 2012 1090 2016 8064 2014 181 Table A-47 Dozers, for lounting on tractors. Cmodity Mm: 723 4052 Equation No. 7 8 9 10 Hodel C C C C CapaC.Util.: .LT.87 0LT067 005067 003067 Dependent ox PX ax PX Constant: “122 “16600 22100 “22400 (“.0132) (“0.935) (1.91) (“1.68) MARUU 0.973 9.38 4.20 “4.75 (0.868) (4.34) (1.13) (“1.11) XRIMF “15.2 “2.53 “37.2 73.4 (“0.818) (“.0710) (“0.784) (1.34) GNP72$ “4.47 “8.98 “14.4 16.7 (“2.30)'I (“2.40)OO (“2.14)*! (2.16)!* PL 5608 “1309 1010 1706 (1.15) (“0.140) (.0177) (0.246) Pg “3052 176 “5305 3502 (“0.148) (3.84) (“0.880) (0.502) D781794 YHCAP83 YHCAP85 YHCAP86 YHCAP87 R“Squared = 034 056 066 068 F= 3.94 10.6 4.27 4.22 df(mn,den) (5,35) (5,33) (5,10) (5,10) Equation Ho. 1 Node) 8 Dependent 0X Constant: 168000 (1.22) UAROU 40.9 (1.83) XRIMF 239 (0.661) GNP72$ “696 (“1060). PL “711 (“00906) Pl “554 (“1028) D781794 40500 (7.41) YHCAP83 YHCAP85 YHCAP86 YHCAP87 "“0 3 00240 (1.90) R-Squared = .60 P: 12.8 df(nul,den) (6,52) DH= 1077 182 Table A-48 Needles, selinq Iachine. Connodity No.: 724 3920 2 8 PX “452 (“3022) “00752 (“3018) “00643 (“1075) 0.115 (2.53)l** 3.33 (4.13) 0.865 (1.95) “3004 (“5.27) 0.158 (1.23) .54 10.3 (6,52) 2.10 0 0X 191000 (1.35) “2035 (“0.187) 95.6 (0.244) “739 (“0.920) “784 (“1.76) 39600 (7.40) “5054 (“2012)*‘ 0.302 (2.43) .60 12.7 (6,52) 1.71 4 D PX “252 (“1.41) “.00729 (“0.467) “00291 (“00602) 2.09 (2.06) 0.592 (1.04) “32.5 (“5001) .00367 (1.16) 0.442 (3.79) .40 5.01 (6,52) 2.35 5 B OX 226000 (1.58) 18.7 (0.740) 28.0 (.0720) “4400 (“0.970) “877 (“8010) “685 (“1.55) 40700 (7.48) “4074 (“1076)** 0.276 (2.20) .61 11.4 (7,51) 6 E PX “466 (“3010) “00577 (“2.04) '00‘52 (“8009) .0941 (8089)! 3.29 (3.90) 0.888 (1.90) “3007 (“5019) .00268 (0.905) 0.202 (1.58) .53 0.12 (7,51) 2.12 183 Table A“49 Needles, seninq Iachine. Connodity No.: 724 3920 Equation No. 7 8 Node) C C Capac.Ut1(.: 0LT065 0LT085 Dependent 0X PX Constant: 176000 “488 (1066) (“1077) UARUU 20.1 “.0140 (1.60) (“0.377) XRIHF “282 “.0523 (“1.78) (0.112) GNP72$ “43.9 .0576 (“2.18).. (0.973) PL “863 3047 (“1071) (2034) PI “68.4 0.389 (“0.244) (0.474) 0781794 67000 “34.6 (29.0) (“5.11) YHCAP83 YHCAPBS YHCAP86 YHCAP87 R“Squared 3 .98 060 F3 176 5.5 «(nu-mm) (6,22) (6,22) 9 10 C C .GE.85. .GE.85 0X PX 132000 “379 (4.47) (“2.12) 50.2 “0.115 (4.70) (“1.78) 146 “1.27 (1.22) (“1.77) “7703 00114 (“4.63)***(1.13) “661 4014 (“3093) (4007) “249 00303 (“1.68) (0.337) 9730 “2005 (5063) (“1096) .93 .69 54.4 8.82 (6,24) (0,20) Equation Ho. 1 Model B Dependent 0X Constant: 271000 (0.724) UARUH 104 (2.17) XRIMF 297 (0.410) CNP72$ “178 (“1076)** PL “2160 (“1026) PX 436 (0.371) 0754794 105000 (7.70) YHCAP83 YHCAP85 YHCAP86 YHCAP87 R-Squared = .87 3 57.5 (if (nu-Men) (6 ,53) 1.97 184 Table A“50 Centrifugal puops for liquids, single-stage, single-suction, close-coupled, under 2-inch. Co-odity ((6.: 742 4026 2 8 PX “3040 (“2020) “00363 (“2005) “00398 (“.0150) 0.523 (1.41)! 14.8 (2.35) 12.4 (2.87) “322 (“6.41) .90 00.0 (6,53) 1.95 3 0 0X 93800 (0.270) 21.3 (0.541) 107 (0.132) “1380 (“0.845) 495 (0.414) 97500 (7.83) “7009 (“1032)§ .86 55.0 (6,53) 1.98 4 D PX “2380 (“1.86) “0.137 (“0.946) 0.121 (.0408) 12.2 (2.04) 12.0 (2.72) “296 (“6.47) .0154 (0.778) .90 77.7 (6,53) 1.83 5 B OX 270000 (0.711) 98.3 (1.26) 260 (0.317) “167 (“1.15) “2130 (“1022) 420 (0.354) 105000 (7.48) “00779 (“0.101) .87 48.3 (7,52) 1.96 6 E PX “3060 (“2020) “00435 (“1052) “00470 (“00156) 0.645 (1.21) 15.1 (2.35) 12.2 (2.80) “325 (“6031) “000900 (“0.319) .90 67.4 (7,52) 1.97 185 Table A“51 Centrifugal pups for )imids, single-stage, single-suction, close-coupled, under 2-inch. Cmodity (40.: 742 4026 Ewation It). 7 8 9 10 Model C C C C Capac.Uti).: 01.1083 01.1083 0GE083 002083 Dependent 0X PX (1X PX Constant: “311000 “1720 4460 “2230 (“0.347) (“0.509) (.00862) (“1.20) W 188 “.0353 11.8 “0.421 (1.34) (“.0669) (0.102) (“1.01) XRIHF 2050 “0.661 “1080 2.36 W728 “137 “0.240 “18.5 0.634 (“0.581) (“0.270) (“.0857) (0.814) P1. “884 12.5 131 7.20 (“0.197) (0.739) (.0523) (0.796) PK 1270 7.82 799 11.4 (0.545) (0.890) (0.421) (1.66) 0754794 92200 “321 801W “260 (3083) (“3054) (3.20) (“2069) W83 W85 W86 W87 R'smim : .92 090 061 091 F: 3M 23.6 21.8 49.2 (If (nu-Men) (6,16) (6 .16) (6 ,M) (6 ,30) Equation No. 11 Rode) Dependent Constant: UARUU XRIHF GNP72$ PL PK 0754794 YHCAP83 YHCAP85 D 0X “272000 (“0.785) 71.3 (1.82) 1230 (1.50) “436 (“0.267) 1090 (0.925) 91500 (7.56) 8.42 (1.47) 56.4 (6,53) 2.00 186 Table A-52 Centrifugal gulps for liquids. single-stage, single-suction, close-coupled, under 2-inch. Connodity No.: 742 4026 12 13 14 E D E 0X 0X 0X 162000 “436000 43600 (0.457) (“1.34) (0.131) 203 96.0 220 (3.52) (2.58) (4.36) 1250 1580 1370 (1.63) (2.12) (2.00) “317 “313 (“2.95)")11 ('3033)“ “1990 286 “1300 (“1.23) (0.182) (“0.856) 616 1140 548 (0.555) (1.03) (0.533) 110000 8710 105000 (8.49) (7.52) (8.82) 16.8 (2.77)lfl8 17.7 24.6 (2.94)III (4.16)388 088 0” 090 56.6 64.0 66.9 (7,52) (6,53) (7,52) 2.25 1.88 2.15 187 Table h“53 Air coapressors, stationary, over 100 hp. Equation No. 1 Node) 8 Dependent 0X Constant: “3360 (“1065) UARO" “00290 (“1020) XRIMF “2.05 (“00366) GNP72$ 1.47 (2.67)830 P1 13.5 (1.29) 9! 6.52 (1.08) 0781794 YHCAP83 YHCAP85 YHCAP86 YHCAP87 Rho 3 00666 (7.29) R-Smared = .27 F: 3096 df(nul,den) (5,53) DU= 2016 mm 9.0.: 743 1035 2 8 PX 67900 (0.749) “1004 (“0.730) 34.1 (0.142) 7.34 (0.267) “333 (“00645) “4601 (“00160) 0.296 (2.38) .10 1.25 (5,53) 1.99 3 0 0X “2900 (“1075) 0.369 (2.55) “9008 (“2005) 12.9 (1.36) 16.1 (30.1) “00395 (“1035)' 0.448 (3.95) .48 9.03 (5,53) 2.11 4 D PX 25800 (0.285) “2007 (“0.260) 145 (0.593) “107 (“00205) “51.5 (“0.178) 3.08 (1.79)** 0.343 (2.81) .14 1.79 (5,53) 2.03 5 8 0X “3180 (“1060) “2600 (“1007) “3091 (“0.746) 1.36 (2.54)Ifl 12.4 (1.20) 8.52 (1.45) “00341 (“00122) 0.640 (6.39) .34 4.50 (6,52) 2.19 6 E PX 18800 (0.195) “4012 (“00272) 154 (0.594) 4.32 (0.152) “6906 (“00129) “5804 (“0.196) 3.09 (1.75)** 0.356 (2.93) .14 1.45 (6,52) 2.04 188 Table A“54 Air coapressors, stationary, over 100 hp. Equation "00 7 lode) C cap3C0Ut110: 0LT066 Dependent 0X Constant: “7120 (“3021) UhROU 0.578 (1.89) XRIMF “4.25 (“0.940) W729 0.319 (0.621) PL 28.1 (2.29) PK 26.2 (4.32) D781794 YHCAP83 YHCAP85 YHCAP86 YHCAP87 R-Squared = .75 F: 1706 Co-odity 1...: 743 1035 8 9 10 C C C .LT.86 .GE.86 .GE.86 PX 0X PX 259000 “3570 “27700 (2071) (“3057) (“2060) 2206 “1002 “5205 (1.72) (“2.27) (“1.18) “155 “1.56 616 (“0.795) (“3.67) (1.46) “6301 1090 122 (“2.86)¢Ifi(2.73)ii (1.76)}! “1220 26.6 “488 (“2030) (6029) (“00796) “258 7.91 “174 (“0.989) (1.59) (“0.353) .31 .80 .41 2.67 14.0 2.48 (5,30) (5,10) (5,13) 189 Table A“55 Air (oppressors, stationary, over 100 hp. Emation Mo. 11 Hode) D Dependent 0X Constant: “2920 (“1.78) UAROU 0.380 (2.69) XRIHF “9.52 (“2014) CNP724 PL 13.5 (1.45) PX 15.9 (3.00) D781794 YHCAP63 “00401 (“1046)5 YHCAP85 YHCAP86 YHCAP87 Rho = 0.453 (3.90) R“Squared 3 049 P= 10.2 df(“ulpden) (5,53) DU= 2005 Cmodity 1b.: 743 1035 12 D PX 36900 (0.398) “4.49 (“0.561) 117 (0.457) “177 (“0.334) “2005 (“00669) 1.99 (1.26) 0.360 (2.96) .12 1.39 (5,53) 2.00 13 14 E E 0X PX “3880 50000 (“2.12) (0.529) “00575 1065 (“2.31) (.0995) “2.79 107 (“0.585) (0.416) 2.15 “12.7 (3.90)““ (“0.371) 13.4 “247 (1040) (“00664) 8.72 3.69 (1.60) (.0122) “00692 2026 (“3.13)!!*(1.19) 0.644 0.339 (6.46) (2.76) .43 .12 6.46 1.19 (6,52) (6,52) 2.07 1.99 Typeuriters, standard, non-portable, electric, nee. Equation Ho. 1 Node) 8 Dependent 0X Constant: “69900 (“00656) UARUU 8.99 (0.800) XRIHP “208 (“0.973) CNP72$ “17.8 (“0.771) PL 842 (1.90) PK 5.98 (.0237) D781794 YHCAP83 YHCAP85 YHCAP86 YHCAP87 Rho = 0.498 (4.41) R-Squared = .18 3 2036 d1(fl“l,d9n) (5053) DB: 2.12 190 Table A-56 Cal-odity N00: 751 1040 2 B PX 1060 (2.36) “000657 (“00963) 1.78 (1.51) “00164 (“1036) “6020 (“2.46) 0.264 (0.187) 0.342 (2.80) .66 20.8 (5,531 1.98 D 0X “122000 (“1075) 2.00 (0.335) “145 (“00759) 1080 (2.74) 23.2 (0.104) 2.11 (1.80)! 0.389 (3.24) .26 3.80 (5,531 2.10 4 D PX 1100 (2.66) “0.114 (“3.18) 1.32 (1.14) “6031 (“2065) “0.174 (“00131) “00167 (“2033)" 0.323 (2.62) .69 24.0 (5,531 1.97 5 B GX “90700 (“1.34) 24.6 (1.88) “218 (“1.19) “6709 (“1096).‘ 993 (2.62) 126 (0.588) 3.63 (2.68)!!! 6 E PX 1120 (2.60) “0.103 (“1.19) 1.31 (1.12) “00233 (“00146) “6037 (“2.62) “00126 (“00932) “.0160 (“1060). 0.323 (2.62) .69 19.6 (6 ,52) 1.96 191 Table A-S7 Typewriters, standard, non-portable, electric, nee. Coo-odity 46.: 751 1040 Equation No. 7 8 9 10 lode) C C C C CapaC.Uti|.: 01.1063 01.1063 060063 005063 Dependent 0X PX 0X PX Constant: 68400 107 “122000 1480 (0.497) (0.115) (“2.06) (4.24) W 2904 “00170 3304 0M9? (1.34) (“1.15) (1.53) (.0460) XRIHF “314 2.90 “67.2 .000359 (“1.51) (2.07) (“0.290) (.00263) W726 “5700 0.300 “4606 “00331 (“1056) (1.22) (“1036)! (“1063) PL 381 “1.98 890 “7.07 (0.518) (“0.399) (2.47) (“3.32) PX “471 .0349 316 0.254 (“1.26) (.0138) (1.07) (0.145) D781794 YHCAP83 YHCAP85 YHCAP86 YHCAP87 R'swam : 062 061 042 066 F: 5065 5029 4054 4706 df(nun,den) (5,17) (5,17) (5,31) (5,31) 192 Table A-58 Radios, household type, lithout phonograph. Equation ”00 1 Node) 8 Dependent 0X Constant: -1400000 (“5.12) UARUU “226 ('4065) XRIHF “1150 ('1060) CNP72$ 590 (6.56).00 PL 8470 (5.23) PK 1070 (1.21) 0781794 0UH743 “7190 ('00300) 0UH781 205000 (8.88) YHCAP83 YHCAP85 YHCAP86 YHCAP87 Rho = “0.173 ('1035) R'SQUITEd 3 085 P: 4201 d11nuI'den) (7,51) DU= 1.98 Co-Odity 1b.: 762 0040 2 8 PX '3309 1'00319) '00104 (“1.17) '00178 ('00633) .0196 (0.626) 0.481 (0.806) .0677 (0.199) “18.0 (‘3065’ '4000 (“0.806) 0.418 (3.54) .24 2.35 (7,51) 1.54 “208000 (“0.361) 93.7 (1.82) 154 (0.103) 826 (0.256) 503 (0.258) 6270 (0.298) 198000 (8.90) '2206 ('2029)*' 0.647 (6.51) .68 15.2 (7,51) 2.16 4 D PX -4005 (“0.372) ‘000475 ('00501) '00120 (-0.426) 0.474 (0.767) 0.121 (0.344) '1703 (‘30067 '3076 (’00775) .00208 (0.983) 0.470 (4.09) .25 2.38 (7,51) 1.50 5 B OX “1270000 ('3095) “242 (”4056) “1440 (’1077) 595 (6.57)¢OI 7950 (4.51) 921 (1.01) “8710 (“0.361) 206000 (8.88) -6040 (0.768) “0.173 (“1.35) 36.. (0,50) 1.95 6 E PX '7904 (“0.712) '00177 ('1011) '00948 1-00328) .0253 (0.815) 0.693 (1.13) .0913 (0.272) '1709 ('3062) ‘4010 (‘00822’ .00228 (1.06) 0.400 (3.36) .26 2.26 (8,50) 1.52 193 Table A“59 Radios, household type, without monograph. Equation "00 7 Node) C CapaC.Uti).: 0LT087 Dependent 0X Constant: -1410000 (-2053) UAROU ’2008 ('3007) XRIHF “1100 (“0.982) W720 561 (4.80)»): P1. 8330 (2.69) PK 1370 (0.948) 0781794 W43 “8350 ('00310) "M781 202000 (7.73) “W83 WBS W86 W87 R-Squared = .82 3 2207 «(...-,8...) (7,36) Co-odity 1b.: 762 0040 8 9 C C .LT.87 .CE.87 PX (IX '8209 '530000 (“0.814) (“0.939) “.0478 “200 (-3.86) (“1.10) 0.266 845 ('1029) (00364) .0726 436 (3.39)"! (1.33) 1.02 5321 (1.80) (1.76) '00237 ‘3330 (-0.896) (“1.12) ‘1902 (’3089) '7021 ('1050) .47 .46 4.54 1.73 (7,36) (5,10) 10 13 £3.87 PX 341 (1.56) 0.130 (1.84) ’00130 (-0.144) '00324 ('2055).' '1041 (‘1020) 0.599 (0.520) .75 6.12 (5,10) Equation HO. 1 Node) 8 Dependent 0X Constant: 1250 (0.438) UAROU .0226 (.0469) XRIMF “11.3 (“1.49) GNP72$ 0.341 (0.377) PL “2.12 ('00127) P“ ‘00148 (“.0161) D781794 YHCAP83 YHCAP85 YHCAP86 YHCAP87 Rho = 0.122 (0.945) R“Squared = 035 F: 5079 df(nun,den) (5,53) DU= 1.98 194 Table 4-60 Shavers, electric. Connodity 80.: 775 4030 2 8 PX “2440 (‘00705) '00184 (’00347) -1309 ('1052) ’1003 1-00997) 24.2 (1.24) 26.5 (2.42) 0.340 (2.77) .42 7.74 (5,53) 1.95 0X 2850 (1.07) '000877 ('00362) '1702 (”2030) -8009 ('00523) 1.36 (0.164) -.0852 ('1045)‘ 0.046 (0.354) .41 7.49 (5,53) 2.00 .0839 (1.29) -0.472 ('1056) '6044 (“00093) 32.2 (1.62) 23.4 (2.12) .0839 (1.29)! 0.365 (3.01) .41 7.32 (5,53) 1.99 5 2 0X 2490 (0.941) '00479 (“0.901) ’1703 (‘2037) 0.861 (0.959) '6048 (“00426) ’00299 (’00360) “0.112 ('1078)** 0.0173 (0.173) .44 6.79 (6,52) 2.02 6 E PX “3670 ('1010) 0.136 (0.246) '1104 (-1027) '1037 (“1036) 30.2 (1.60) 27.9 (2.69) 0.101 (1.56)! 0.290 (2.33) .48 8.16 (6,52) 1.96 Equation No. 7 Node) C Capa£.Uti|.: 011066 Dependent 0X Constant: 5520 (1.30) “9RD” '00333 ('00566) XRIHF “11.9 (‘1036) GNP72$ 0.959 (0.972) PL “28.0 (“1.18) PK ‘9084 (“0.846) 0781794 YHCAP83 YHCAP85 YHCAP86 YHCAP87 R'SQUBTEd : 044 F: 4060 «(...-,6...) (5,30) 195 Table 9'61 Shavers, E)ECtT1C0 Min; No.: 775 4030 8 9 10 C C C .LT.86 .GE.86 .GE.86 PX 0X PX 116 2310 “3370 (00275) (00667) ('1002) 00968 00195 '1062 (1066) (00125) (-1009) -1002 '2309 '1600 (“1.20) (“1.62) (“1.29) '2038 '1016 00761 (“2.44)Il (“0.482) (0.331) 4.50 11.0 50.3 (0.193) (0.513) (2.47) 2206 '1012 7071 (1.96) (“.0652) (0.471) .30 .38 .77 2.51 2.18 11.8 (5,30) (5,18) (5,18) Equation 80. 1 Hodel B Dependent 0X Constant: “12000 (“2.14) UAROU 0.482 (0.657) XRIHF “3.10 ('00210) (0P72$ 290 (1.86)! PL 65.5 (2.18) PX 7.93 (0.463) D781794 YHCAP83 YHCAP85 YHCAP86 YHCAP87 Rho 3 00572 (5.35) R'SQUIPEd 3 056 P= 13.7 df(nul,den) (5,53) DU: 2029 196 Table A-62 Uacuun cleaners, electric, household type. Co-odity )(o.: 775 7520 PX 7740 (0.838) 1.28 (1.30) 21.7 (0.925) '1061 (’00634) ’5004 ('1013) 1.78 (.0690) 0.833 (11.6) .07 0.836 (5,53) 2.06 3 D 0X “1880 (”00334) 1.13 (2.24) ‘1607 ('1014) 29.7 (0.947) '7010 ('00383) '00138 ('1052)* 0.709 (7.72) .38 6.64 (5,53) 2.56 4 D PX 5030 (0.647) 0.916 (1.17) 28.0 (1.33) -4407 ('1004) '00534 (“0.437) '00534 (“0.437) 0.843 (12.0) .07 0.812 (5,53) 2.02 5 E GX “11000 ("2002) 0.154 (0.206) '8004 ('00556) 3.24 (2.13).! 60.7 (2.07) 9.61 (0.580) “00149 (”1071)'* 0.548 (5.03) .61 13.6 (6,52) 2.36 6 E PX 7960 (0.852) 1.25 (1.26) 21.5 (0.904) '1053 (“00592) ’5204 (’1016) 1.93 (.0740) '00464 ('00376) 0.836 (11.7) .08 0.713 (6,52) 2.02 197 Table 8-63 Uacuua cleaners, electric, household type. Connodity 80.: 775 7520 Equation No. 7 8 9 10 lode) C C C C Capac 0111-1103 01.1085 01.1065 0W085 £8.85 Dependent 0X PX 8X PX Constant: “14200 “187 “13000 -6780 (“1.70) (“.0166) (“4.89) (“0.858) W 1096 3047 -1059 ’2056 (1.76) (2.32) (“1.63) (“0.882) XRIHF “11.2 17.3 “6.95 “143 (“0.798) (0.912) (“0.670) (“4.65) GNP72$ 1.64 “4.43 5.61 “1.76 (0.935) (“1.87)I (3.79)018 (“0.400) P1. 6205 '3052 7503 130 (1.84) (“.0585) (4.87) (2.83) PX 9.10 23.0 3.91 103 (0.368) (0.690) (0.294) (2.61) 0781794 YHCAP83 YHCAP85 YHCAP86 YHCAP87 R'smm = 089 026 091 069 P: 3702 1056 5305 1104 df(nu,den) (5,23) (5,23) (5,25) (5,25) 198 Table 8-64 Vacuuo cleaners, electric, household type. Coo-Odin} 1b.: 775 7520 Emation lb. 11 12 13 14 Ibdel D D E E Dependent 0X PX 0X PX Constant: “3160 6360 “9870 4980 (“0.557) (0.646) (“1.65) (0.535) W 1.20 0.828 0.729 0.667 (2.30) (1.07) (0.007) (0.655) XRIHF ’6047 2303 '1079 2601 (“0.565) (1.14) (“0.115) (1.12) W729 2.16 0.746 (1.19) (0.264) P1. 40.1 “54.7 59.0 “51.5 (1.27) (‘10327 (1069) (‘10161 PK “16.1 7.26 “0.520 7.71 (‘006537 (00268) (‘002907 (0.3)2) 0781794 W83 00137 '00222 00734 ’00237 (1.60) (“1.99)“ (0.782) (“1.90). MOS WM WW 3 0.754 0.852 0.639 0.856 (8.82) (12.5) (6.38) (12.7) R“Smared ‘ 032 013 0‘9 014 F: 5010 1064 8.29 1037 dflmmden) (5,53) (5,53) (6,52) (6,52) “3 2043 2002 2032 2002 199 Table A-6S vacuul cleaners, electric, household type. lode) C hWCOUt110: 011063 Dependent 0X Constant: “8630 (“0.845) UARUU 3.08 (1.90) XRIHF “8.24 (’0053‘4) GNP72$ .0425 (.0157) PL 57.5 (1.05) PK '1109 (“0.429) 0781794 YHCAP83 YHCAP85 YHCAP86 YHCAP87 R“SQRN(1 3 068 F: 2506 df(nul,den) (5,17) Cmodity 110.: 775 7520 16 17 18 C C C .LT.83 .GE.83 .GE.83 PX 0X PX “433 “19400 “10200 (“.0306) (“6046) ('1059) 4019 '1002 '1094 (1.86) (“0.921) (“0.824) 502 6096 “142 (1.18) (0.595) (“5.67) '6000 6006 ’2028 (“1.61) (3.50)Olt (“0.613) “15.2 95.6 144 (“0.201) (5.25) (3.69) 40.0 15.2 115 (1.04) (1.01) (3.58) .28 .92 .71 1.30 67.8 15.2 (5,17) (5,31) (5,31) 200 Table A-66 Toasters, autonatic, electric, household type. Emation No. 1 Hodel B Dependent 0X Constant: “2560 ('2009) UARUU .0486 (0.398) XRIHP '2045 (“1.05) CNP72$ 0.616 (2.45)”! PL 11.0 (2.27) PK 6.05 (2.21) D781794 YHCAP83 YHCAP85 YHCAP86 YHCAP87 Rho = 0.498 (4.42) R-Squared = .65 P: 1909 df(nul,den) (5,53) DU= 1.89 Cal-odity N00: 775 6625 2 B PX 8560 (0.417) ‘3056 ('1006) “101 ('1087) ’0066 (“1.40) 174 (1.48) 18.9 (0.289) 0.236 (1.86) .55 12.8 (5,53) 1.95 0X “1370 (’1044) 0.300 (3.60) '3061 (“1.53) 6.13 (1.14) 5.53 (1.78) .0371 (0.235) 0.619 (6.06) .49 10.2 (5,53) 1.88 PX 13400 (0.679) -9034 (.5022) “130 (“2.35) 162 (1.44) “5025 (’00656) '00778 (-2007).' 0.200 (1.57) .58 14.9 (5,53) 1.97 5 B OX “2570 ('2063) .0519 (0.400) '2040 ('1000) 0.613 (2.39)OI 11.0 (2.24) 6.06 (2.19) .00124 (.0828) 0.498 (4.41) .65 16.3 (6,52) 1.89 6 E PX 16600 (0.809) ‘6077 (’1079) “134 (’2036) ’5010 (“0.781) 150 (1.30) 5.21 (.0813) '00666 (“1.65)U 0.216 (1.70) .58 12.0 (6,52) 1.97 201 Table A-67 Toasters, autonatic, electric, household type. Co-odity 110.: 775 0125 Equation No. 7 8 9 10 Hodel C C C C CapiC.Ut1).: 01.1085 01.1085 0E085 £8.85 Dependent ox PX ox rx Constant: “4140 19100 “2570 1160 (“2.43) (0.518) (“4.49) (.0649) WROU ‘00855 ’7040 ‘00896 00546 (“0.378) (“1.51) (“0.425) (.0828) XRIHF “5.53 “198 “4.58 “14.6 (“1.93) (“3.19) (“2.05) (“0.209) GNP72$ 0.850 “8.75 0.642 “6.31 (2.38)¢fl (“1.13) (2.01)! (“0.632) PL 21.4 257 14.7 “36.2 (2.35) (1.30) (4.40) (“0.347) PK 9.84 “39.4 5.08 171 (1.96) (“0.361) (1.77) (1.91) 0781794 YHCAP83 YHCAP85 YHCAP86 YHCAP87 R“Sq1ared 3 063 072 090 076 P: 2208 1106 0501 1600 «(...-,1...) (5,23) (5,23) (5,25) (5,25) 202 Table A“68 Sun or glare glasses, and sun goggles. Equation No. 1 Hodel B Dependent 0X Constant: “11100 (”1012) UhRUU 4.46 (3.05) XRIMF 27.4 (0.978) W724 “0.616 (“0.188) PL '4023 (“.0763) PR 49.4 (1.65) 0781794 “1490 ('3079) DUH751 998 (2.14) DUH752 969 (2.05) DUH764 3050 (7.81) YHCAP85 Rho 3 00510 (4.56) R-Squared = .74 3 1508 df(nul,den) (9,49) = 2001 Cmodity 11).: 304 2220 2 8 PX 11800 (1.01) ‘002‘9 1'00361) ‘2405 (“1.41) “0.592 (‘00206) 8.41 (0.276) 11.8 (0.671) 808 (3.38) “1010 ('4060) 464 (2.25) “241 (“1.52) 0.994 (4.0.0) .05 10.0 (9,49) 1.47 GX “14400 (“1.61) 4.40 (5.45) 30.4 (1.16) 9.81 (0.184) 54.7 (1.94) “1560 ('4056) 1020 (2.21) 991 (2.20) 3080 (7.83) .0944 (0.604) 0.463 (4.02) .76 16.8 (9,49) 2.01 PX 9360 (1.38) “00301 1‘00444) “2106 ('1031) 15.7 (0.577) 10.6 (0.611) 802 (3.42) -994 (‘50307 482 (2.61) “240 (“1.55) .0991 (1.30)! 0.994 (4.7.4) .04 10.4. (9,49) 1.37 0X “13600 (“1.39) 4.70 (3.05) 28.3 (1.01) '00734 ('00223) 7.12 (0.129) 56.2 (1.93) “1520 (”3092) 998 (2.10) 954 (1.98) 3060 (7.63) .0976 (0.617) 0.462 (4.00) .70 14.3 (10,411) 2.02 6 E PX 14200 (1.24) ’00268 1.00391) '2402 (‘10407 ”1062 (“0.550) 8.72 (0.289) 9.85 (0.563) 790 (3.33) “1040 ("0987 434 (2.11) “252 1'10607 0.111 (1.39)! 0.993 (66.7) .00 9.39 (10,413) 1.34 203 Table A-69 Sun or glare glasses, and sun goggles. Equation No. 7 Hodel C Capac.Util.: 0L1065 Dependent 0X Constant: “25600 ('1088) UhRUU 6.66 (3.24) XRIHF 47.3 (1.49) CNP72$ 0.216 ('00595) PL 45.3 (0.563) PK 72.4 (1.65) 0781794 “1830 ('4045) 0UH751 1130 (1.99) 0UH752 981 (1.75) DUH764 3220 (7.30) YHCAP83 YHCAP85 YHCAP86 YHC8P87 R“Squared 3 093 = 3000 «(...-,1...) (9,19) Connodity No.: 8 9 C C .LT.85 .CE.85 PX 0X “17300 “3570 (“0.960) (“0.322) “1.53 6.23 (“0.562) (2.08) “79.8 54.1 ('1069) (1069) 2022 '6002 (00462) ('1013) 158 '9015 (1.48) (0.159) 42.8 3.33 (00733) ('00763) 1580 “744 (2090) ('1023) “1500 (“1.98) 252 (0.338) “486 (“0.830) .83 .60 10.5 5.96 (9,19) (0,24) 884 2220 10 C .GE.85 PX 3890 (0.934) 1.28 (1.14) ‘6097 ('00576) ‘2061 ('1030) 286 (0.132) -1204 (“0.775) 2840 (12.5) .98 203 (6,24) Equation No. 11 Hodel Dependent Constant: WWW ”(INF W720 Pl. PX 0781794 0111751 0111752 W64 Table A-70 a... or glare glasses, and sun goggles. Co-odity 1b.: 884 2220 12 13 14 D D E E (1X PX 0X PX -13300 9860 “12600 8010 (“1.46) (1.46) (“1.5) (0.678) 4.39 .0264 4.58 .0166 (5.38) (.0386) (3.08) (.0241) 3209 'fi06 3101 '3400 (1.25) (“2.05) (1.08) (“1.95) “00521 00660 (“0.158) (0.228) 5.06 2.36 2.97 5.16 (.0932) (.0881) (.0525) (0.174) 48.4 13.4 49.5 13.7 (1.66) (0.792) (1.65) (0.798) “1540 808 “1520 812 (“4.38) (3.52) (“3.83) (3.49) 1030 “999 1020 “979 (2.27) (“5.43) (2.16) (“4.77) 1010 448 985 467 (2.28) (2.46) (2.06) (2.32) 3050 “239 3040 “234 (7.90) (“1.57) (7.71) (“1.51) .0971 “0.136 .0961 “0.140 (00630) ("1069). (00617) ('1067)' 0.502 0.994 0.502 0.994 (4.46) (68.9) (4.46) (70.3) .75 .67 .75 .67 16.1 11.2 14.2 9.86 (9,49) (9,49) (10,48) (10,48) 10% 1026 1098 1026 204 205 Table A-71 Sun or glare glasses, and sun goggles. Col-odity No.: 884 2220 Equation No. 15 16 17 18 Hodel E E E E Dependent 0X PX 0X PX Constant: “10000 24300 “10800 “3320 ('00949) (1061) (”1003) ('00582) URRUU 5.68 “0.360 4.73 0.162 (3.51) (“0.683) (3.12) (0.344) XRIHF 50.1 “16.2 3.82 “26.9 (1.66) (“1.00) (1.26) (“1.68) GNP724 “1.95 “2.17 “0.584 4.16 (“0.552) (“0.741) (“0.171) (1.89)! P1. “22.1 6090 '1007 1309 (“0.381) (2.63) (“0.181) (0.567) PX 40.3 4.06 42.2 2.08 (1.27) (0.249) (1.34) (1.32) 0781794 “1400 610 “1490 752 (“3.48) (2.78) (“3.66) (2.67) DUH751 1160 “1080 1070 “914 (2.49) (“6.07) (2.27) (“5.66) DUH752 1040 398 1030 491 (2.22) (2.28) (2.14) (3.03) DUH764 3100 “205 3050 “169 (7.71) (“1.66) (7.69) (“1.48) YHCAP85 0.144 0.141 YHCAP86 .0827 “0.214 (0.530) (“3.68)030 Rho 1 = 0.416 1.36 0.505 1.41 (3.22) (11.2) (3.85) (12.2) ”)0 2 = 00179 ’00366 00413 '00471 (1.38) (“3.00) (0.314) ('4.06) R'smam 3 073 076 074 000 = 1300 1405 1303 1003 df(nul,den) (10,47) (10,47) (10,47) (10,47) 3 2003 2009 2.00 203 206 Table A-72 c1OCKS, EleCtr1C0 Con-odity No.: 885 2020 Equation No. 1 2 Node) 8 B Dependent 0X PX Constant: “1000000 “44300 (“0.970) (“2.99) VAROU 394 “3.82 (2099) 1'1074) XRIHP 836 19.5 (0.309) (.0499) CNP72$ “248 9.95 (“0.874) (2.29)** PL 1650 2.40 (0.302) (2.90) PK 5600 99.6 (1.79) (2.14) 0781794 YHCAP83 YHCAP85 YHCAP86 YHCAP87 Rho 3 00600 00396 (5.77) (3.31) R-Squared = .28 .24 P: 4021 3023 df‘flUI,dEN) 15,53) (5,53) D“: 1090 2011 3 0 8X “1430000 ('1046) 322 (3.74) 2760 (1.06) (0.468) 5120 (1.58) '1034 ('00883) 0.653 (6.63) .25 3.50 (5,53) 1.95 4 D PX “38500 (‘2062) 1.10 (0.869) 16.4 (0.407) 220 (2.62) 114 (2.40) 0.487 (1.98)** 0.416 (3.52) .21 2.50 (5,53) 2.08 5 E OX “960000 (“0.965) 424 (2.01) 179 (.0679) “334 (”1007) 1940 (0.362) 5940 (1.94) 8.75 (0.524) 6 E PX “44600 1’2098) '2031 (“0.850) 27.1 (0.673) 7.36 (1.42) 240 (2.87) 103 (2.18) 0.272 (0.949) 0.407 (3.43) .24 2.79 (5,52) 2.09 207 Table A“73 CIOCKS, e1ECtT1C0 Co-odity n... as 2020 Equation No. 7 8 flodel CapiC.Util.: 011083 C C .LT.83 Dependent 0X PX Constant: “ARCH XRIMF GNP72$ PL PK D781794 YHCAP83 YHCAP85 YHCAP86 YHCAP87 R-Squared = f: dflmfien) (5,17) “634000 “57200 (‘003617 (-1069) 884 5.58 (3.16) (1.04) “1430 137 (“00540) 12069) '96? 00264 (“2.08)ll (.0294) 3700 228 (0.394) (1.27) 3062 114 (0.642) (1.24) .82 .41 15.4 2.39 (5,17) 9 10 C C .CE.83 .GE.83 0X PX “772000 “51400 ('1014) 4‘4038) 9600 ’3063 (0.384) (“0.839) “775 “112 (“0.292) (“2.45) 206 2.85 (0.522) (0.419) 1640 404 (0.397) (5.65) 3800 154 (1.12) (2.62) .62 .64 10.0 10.8 (5,31) (5,31) 208 Table A“74 Clocks, electric. Cpl-odity "003 685 2020 Equation (40. 11 12 Hodel C C Capac.Util.: .LT.85 .LT.85 Dependent 0X PX Constant: 268000 “57000 (0.176) (“2.29) UARUH 762 1.08 (3.75) (0.327) XRIMF “2650 105 (“1.03) (2.50) CNP72$ “955 7.87 (“2.98)30¢(1.50) PL “2090 249 (“0.256) (1.87) P! 4010 93.2 (0.888) (1.27) 8781794 YHCAP83 YHCAP85 YHCAP86 YHCAP87 Rho = R'SQU3FEd 3 073 045 P= 12.5 3.71 df(nul,den) (5,23) (5,23) m: 13 C OCEOBS 0X “1840000 (‘2078) 125 (0.514) 2410 (0.934) 395 (1.07) 5880 (1.53) 3740 (1.13) .76 15.5 (5,25) 14 C .GE.85 PX “40500 ('2090) ’4000 ('00934) “118 (“2.16) 2.45 (0.315) 377 (4.63) 110 (1.58) .66 9.37 (5,25) 15 E 8X “1400000 (’1068) 469 (3.54) “838 (“0.375) “389 ('1058) 4560 (0.988) 7520 (2.92) 35.9 (2.42).! 0.382 (3.18) .49 8.35 (6,52) 1.89 16 E PX 43500 (“2.86) '3094 (“1.64) 18.2 (0.442) 10.1 (2.24)8fl 238 (2.81) 98.8 (2.09) “00353 ('00130) 0.399 (3.34) .24 2.67 (6,52) 2.11 209 Table A-75 Tape, mane-sensitive, plastic. Co-odity 1b.: 891 0945 Emation lb. 1 2 3 4 5 6 lbdel 8 8 D I) E E Dependent 0X PX 0X PX 0X PX Constant: “69800 4870 15500 881 “90200 3150 (“0.781) (1.08) (0.203) (0.208) (“1.10) (0.695) m 18.6 “1.14 27.9 “1.42 17.4 “0.907 (1099) ('1066) (3046) ('3083) (1065) ('1051) XRIHF “109 “6.82 “249 5.72 “33.7 “0.881 (“0.482) (“0.578) (“1.15) (0.511) (“0.145) (“.0732) W724 45.9 “1.44 47.6 “1.46 (1.82)! (“1.14) (1.94)! (“1.17) P1. 3.90 4.60 “207 18.7 106 12.2 (.00915) (0.189) (0.498) (0.776) (0.246) (0.501) PK 7407 '4016 4406 ’6069 7‘07 '5039 (0.304) (“0.303) (0.177) (“0.498) (0.303) (“0.394) D781794 W83 W85 W86 0.898 0.134 1.10 0.132 (0.824) (1.92)“ (1.03) (1.90)“ W87 Rho 3 00663 00531 00677 00569 00642 00549 (13.1) (4.82) (14.0) (5.31) (12.0) (5.05) R“Squared = 032 056 026 056 036 059 P: 0096 1500 3063 1302 4090 12.5 «(...-,4...) (5,53) (5,53) (5,53) (5,53) (6,52) (6,52) DH= 2.10 2.02 1.97 1.98 2.05 1.97 210 Table A-76 Tape, pressure-sensitive, plastic. Conoodity u... 091 0945 Equation Ho. 7 8 9 10 Hodel C C C C Capac.Util.: 01.1086 01.1066 00E066 0CE066 Dependent 0X PX 0X PX Constant: “308000 6040 “309000 6320 (“3.65) (1.70) (“3.93) (1.43) m '1105 -1060 -4201 "00625 (“0.986) (“3.42) (“1.19) (“0.314) XRIHF 82.2 0.555 “740 “57.6 (0.478) (.0764) (“2.21)44 (“3.06)"4 CNP72$ 104 0.127 129 “5.91 (5.34) (0.154) (2.34) (“1.91) PL 1250 “8.24 1390 66.7 (2.67) (“0.418) (2.85) (2.43) PX 363 “15.9 973 “5.23 (1.58) (“1.64) (2.48) (“0.238) D781794 YHCAP83 YHCAP85 YHCAP86 YHCAP87 R“Smared : 068 063 064 090 F: 4206 $01 1906 33.9 «(m-,den) (5,30) (5,30) (5,10) (5,10) Equation Mo. 1 Hodel 8 Dependent 8X Constant: 4410 (0.595) UAROU 2.60 (3.16) XRIHF 3.82 (0.195) CNP720 “0.316 (“0.149) PL ’2502 ('00692) PK -2906 ('1035) D781794 “1890 (‘6063) DUH742 2090 (9.67) YHCAP83 YHCAP85 YHCAP86 YHCAP87 3 00620 (11.0) R’SQHIPEd 3 076 3 2507 «(...-,0...) (7,51) 003 2042 Pens, ball-point type. Coauodity Ho.: 895 2115 2 8 PX “7480 (‘00591) 0.581 (0.268) 38.1 (1.22) 3.51 (0.777) 21.2 (0.287) '2011 (”00579) 6940 (12.9) “1600 (‘2043) 0.263 (2.09) .92 84.5 (7,51) 2.00 211 Table A“77 0X 5020 (0.800) 2.49 (4.01) 1.92 (0.111) -3302 (“0.970) '2509 ('1019) “1860 (“6.69) 2100 (10.2) ‘00157 (“1.71)!! 0.842 (12.0) .79 20.0 (7,51) 2.34 4 5 D E PX GX “10100 2350 (“0.817) (0.314) 2.57 2.13 (2.43) (0.397) 39.0 7.76 (1.16) (0.397) 1.60 (0.686) 3509 '2606 (0.492) (“0.754) 14.4 “26.5 (0.378) (“1.21) 6900 “1900 (14.0) (“6.71) “1720 2130 (“2.68) (10.1) 0.273 “0.188 (1.21) (“1.85)§l 0.325 0.851 (2.64) (12.4) .91 .80 74.2 24.5 (7,51) (0,50) 2.04 2.35 6 E PX “11100 (”00636) 1.95 (0.733) 39.8 (1.18) 1.32 (0.254) 40.4 (0.527) 11.7 (0.300) 6840 (12.3) “1690 ('2055) 0.240 (0.930) 0.321 (2.61) .91 04.3 (0,50) 2.04 Equation No. 7 Node] C Capac.Util.: 0LT083 Dependent 0X Constant: “31600 ('5005) UfiRUU 1.30 (1.19) XRIHF “15.9 ('8066) CNP72$ 4.34 (2.28).” PL 177 (5.22) PK 40.5 (2.35) 0781794 “1820 ('9007) DUH742 YHCAP83 YHCAP85 YHCAP86 YHCAP87 R“Squared 3 096 F: 62.2 (um-,0...) (0,10) 212 Table A“78 Pens, ball-point type. Co-odity ((0.: 095 2115 8 C .LT.83 PX 32100 (1.38) ‘00301 (’00742) 5.17 (0.145) -1092 ("00278) “196 ('1055) '1208 ('00889) 6720 (9.01) .88 20.2 (0,10) 9 10 C C .GE.83 .GE.83 0X PX 7960 “13800 (1030) ('8011) 4005 ’1003 (2.13) (“0.267) '4901 7209 (“2.50) (1.84) “8.54 9.23 (“2.59)*!*(1.38)! 7.41 25.8 (0.205) (0.352) 11.5 “18.9 (0.474) (“0.384) “716 6760 (‘2036’ (1100) 2150 “1690 (S095) (-2031) .83 .97 20.7 146 (7,29) (7,29) Equation "00 11 Node) 0 Dependent 0X Constant: 4290 (0.666) UARUU 2.53 (4.06) XRIMF 4.46 (0.254) GNP72$ PL '2601 (“0.746) PK '3005 (“1.38) 0781794 “1910 (“6.82) 0UH742 2100 (9.96) YHCAP83 YHCAP85 “.0438 (“0.444) YHCAP86 YHCAP87 3 00.29 (11.4) R“Squared 073 3 2600 (If (II-,den) (7 ,51) D“: 2039 213 Table A“79 FEDS, b311’p01nt ty990 Co-odity No.: 095 2115 12 D PX “14000 (‘10097 2.77 (2.49) 37.8 (1.09) 56.6 (0.754) 24.0 (0.606) 6940 (14.5) “1740 (’2077) 0.367 (1.59)! 0.378 (3.14) .90 00.3 (7,51) 2.03 13 E 0X 4660 (0.620) 2.58 (3.11) 3.55 (0.178) '00222 ('00803) '2702 (“0.735) -3002 ('1035) “1900 (“6.59) 2090 (9.63) '00426 (“00427) 0.828 (11.3) .78 22.3 (8,50) 2.39 14 E PX “17300 4'1023) 1.67 (0.727) 40.6 (1.14) 2.56 (0.550) 71.5 (0.898) 21.0 (0.516) 6770 (12.0) “1680 ('2064) 0.350 (1.48)! 0.383 (3.19) .90 56.6 (8,50) 2.02 214 Table A-80 Pens, ball-point type. mm 111.: 095 2115 Equation Ho. 15 16 17 18 Hodel C C C C CEPBC0Util0: 0LT085 0LT085 0C£085 062085 Dependent 0x PX 0x 9x Constant: “21000 17400 22600 “36600 (“3.52) (0.910) (3.20) (“2.71) UQRUU 2.58 “0.516 4.40 “3.02 (2.95) (“0.184) (2.24) (“0.807) XRIHF “12.8 14.8 “31.9 46.6 (“1.27) (0.457) (“1.57) (1.21) GNP72$ 1.21 “0.254 “11.8 16.4 (0.794) (“.0520) ('3.42)I§§(2.49)I* PL 117 “114 “43.2 118 (3.58) (“1.09) (“1.16) (1.67) PK 2809 3043 ’4500 4809 (1.58) (.0585) (“1.65) (0.942) 0781794 “1780 7410 52.1 5970 (“9.95) (12.9) (0.135) (8.10) DUH742 2230 “1770 (6084) 4'2085) YHCAP83 YHCAP85 YHCAP86 YHCAP87 R’smat‘ed 3 093 096 0 89 098 2: (0.0 01.0 20.7 140 df(m,den) (0,22) (0,22) (7,23) (7,23) BIBLIOGRAPHY Bibliography Adams, F. G. , Eguchi, H., & Meyer— zu- -Schlochtern, F. 53 Econometric Ana1ys1s of International Trade. Paris: 0. E. c. D. 1969. Adams, F. G., 6 Junz, Helen B. "The Effect of the Business Cycle on Trade Flows of Industrial Countries, " Journa1 of Finance, Vol. XXVI, No. 2 (May 1971), pp. 251- 268 Artus, Jacques R. "The Short- Run Effects of Domestic Demand Pressure on British EXport Performance" IMF Staff Papegg, Vol. 17, No.2 (July 1970), pp. 2T7“ 276. Artus, Jacques R. & Rhomberg, Rudolf R. "A Multilateral Exchange Rate Model," IMF Staff Papers, Vol. 20, No. 3 (November 1973), pp. 591- 611. Ball, R J. "Credit Restriction and the Supply of EXports, " The Manchester School of Economic and Social Studies, 29 1961, pp. 161-172. “‘ ““ < 0 Ball, R. J., Eaton, J. R., & Steuer, M. D. "The Relationship Between U.K. EXport Performance in Manufactures and the Internal Pressure of Demand," Ecoggm1c iQBCQEl) Vol. 76, September 1966, pp. 501- 518. Batchelor, R. A., & Bowe, C. "Forecasting U.K. International Trade: A General Equilibrium Approach," App11§q Sggngm1cS, Vol. 6, 1970, pp. 109-141. Bautista, Romeo M. "Effects of Major Currency Realignment on Philippine Merchandise Trade, " Review of Economics and ...-“..-- ~——- Statistics, Vol.59, No. 2 (May 1977), pp. 152- 160. --.---..— Bergstrom, A. R. "An Econometric Study of Supply and Demand for New Zealand's EXports," Econometr1ga, Vol. 23 (July 1955), pp. 258-276. Brechling, Frank, & Wolfe, J. N., "The End of Stop-Go," Lloydg Sang Rgg1ew, No. 75 (January 1965), pp. 23-30. Clark, Peter K. "Potential GNP in the United States, 1908- 80," Review 91 1gggme 33g flea11h, June 1979, pp. 141- 165. COOper, R. A., Hartley, K., & Harvey, C. R. M. EXQort Performance and the Pressure of Demand: A Study of u.- .... —.-.—.---.‘ —.—— Firms'f‘fa'fidofi?~ Cé6rgé‘Kllen and Unwin, 1970. 215 216 De Leeuw, Frank; Forest, Lawrence R. Jr.; Raddock, Richard D.; & Kenesey, Zoltan E. "Measurement of Capacity Utilization: Problems and Tasks," Staff Studies, .- ...—“<— Board of Governors of the Federal Reserve System, July 1979. Dean, Joel. Statistical Determination of Costs with Special —--—..——._-- ~.—~—-—-—-.—-—.--—_— .—.—- “--—— “-.. .— ..--.-‘ Reference to Magg1na1 CosES. Chicago: University of fl-S‘- “ Chicato Press, 1936. Dean, Joel. Statistical Cost Estimation. Bloomington: ---.‘—._— ——-—-—- --—.—-_._——_._.— Dixit, A. K. "The Balance of Trade in a Model of Temporary Equilibrium with Rationing," Review of gggnom1g Stud1es, Vol. 45, No. 141 (Oct. 1978), pp. 393-HOH. Dunlevy, James A. "A Test of the Capacity Pressure Hypothesis Within a Simultaneous Equations Model of EXport Performance," Review of Economics and Stat1§§1c§, Vol. 62, N6. T'(FEDT‘T9667I‘pfif‘131-135. Geraci, Vincent, & Prewo, Wilfried. "Bilateral Trade Flows and Transport Costs," Review of Economics and ‘1‘ ‘. Stati§t1g§, Vol. 59, No. 1 (Feb.“197777‘5p?‘67-7u. Goldstein, Morris, & Khan, Mohsin S. "The Supply and Demand for EXports: A Simultaneous Approach," Reviey of Economics and Sta§1§§1g§, Vol. 60, No. 2 (May 1978), pp. 275-286T‘ _“ Goldstein, Morris, & Officer, Lawrence H. "New Measures of Prices and Productivity for Tradable and Nontradable Goods," Revieg of 1ncgme and Wealth, Vol. 25, No. u (Dec. 1979), pp. H13-427. Gordon, R. J. "A Comment on the Perloff and Wachter Paper," CarnegieRochester Conference Series on 399113 Policy, --..«—-—«-.- Vol. 10, 1979, PP. 187-19M. Greene, Margaret L. Waiting 11me1 A Eagtggl1n EXQort Demand £95 Manufiagtuggs. Princeton Studies in International Finance, No. 37. Princeton: International Finance Section, April 1975. Hadar, Josef. Elementa: Iheggy 91 fl1croeggggm1g Behavior (Second 56.). REa ing, ass.: Addison-Weleyf”T97H. Henderson, James M., & Quandt, Richard E. Microeconomic Theory; A Maghgmatical Approach (Rev. Edi). New York: McGraw Hill, 1971. 217 Henry, George B. "Domestic Demand Pressure and ShortRun EXport Fluctuations," Yale geeggmte Essays, Spring 1970, pp. "3-81. International Monetary Fund. Integeettggel theggtel Statistice, Vol. 33, No. 3 (March 19807. Johnston, John. Statistical Cost Analysis. New York: ———--—————- -“4- “..-.- ———_ McGraw-Hill, 1960. Klein, Lawrence R. "Some Theoretical Issues in the Measurement of Capacity," Econemetttee, Vol. 28, No. 2 (April 1960), pp. 272-286. Klein, Lawrence R. & Long, Virginia. "Capacity Utilization: Concept, Measurement and Recent Estimates," Brooktggs Pagers 99 Eeeaemie Aeiixiix. 1973. pp. 45-56. Klein, L. R. & Summers, R. The Wharton legeg of gepeeity Utilization. Philadelphia: University of Pennsylvania FFééé: 1966. Kmenta, Jan. Elements of Econometrics. New York: Macmillan, 1971. “““““““““““ Kreinin, Mordechai E. "Disaggregated Import Demand Functions- Further Results," 30929259 EEQQQEEE lgglflél: V01- “0: No. 1 (July 1973), pp. 19-25. Kreinin, Mordechai B. International Economics: A Policy -----__.~—- _._.—--_---—.~ --.-..- Apptgeeh (Second _EETY.7 New York: Harcourt Brace Jovanovich, 1975. Kreinin, Mordechai E- lflfiECEEEZQQEL .SQQQQiEii A EQLEEX Appreeeh (Third Ed.). New York° Harcourt Brace Jovanovich, 1979. Lancaster, Kelvin J. "A New Approach to Consumer Theory," Journet gt Petttteel Econemy, Vol. 74 (April 1966), pp. 132-T57. Mintz, Ilse. Cyclical Flggtgatioge 13 the gtpette of the United Steteg-§tfiee 1§19. New York: National Bureau 0? Economic Research, 1967. Morgan, D. J., & Corlett, w. J. "The Influence of Price in International Trade: A Study in Method," Journel 92 the Reyet Stettstical Society, Series A, VoI. 115, Part III, 1951, pp. 307-352. 218 Muellbauer, John, & Winter, David. "Unemployment, Employment, and EXports in British Manufacturing: A Non-Clearing Markets Approach," Manuscript presented at International Seminar on Macroeconomics, Paris, September 10-11, 1979. Nurske, Ragnar. "The Relation Between Home Investment and External Balance in the Light of British EXperience, 1945-1955," Review of Economics and Statistics, Vol. 38 _-— “..-—...“- (May 1956), pp. 121-154. Officer, Lawrence H. "Effective Exchange Rates and Price Ratios over the Long Run: A Test of the Purchasing- Power- -Parity Theory," Canadian Journal of Economies, Vol. 13, No. 2 (May 1980 0), pp. 206-230. " -_“ Okun, Arthur M. "Potential GNP: Its Measurement and Significance, " in American Statistical Association, Proceedings of the Business and Economies Section, ———-—-—--—-—¢.- ---- 1962, pp 98-1ou. Organization for Economic COOperation and Development. Main Economic Indicators HistoLieal Statistics 1960-1975. ..-“ ---—-r—-- --—-_-——-.—--— ...—-----— Paris, Oct. 1976. Pagoulates, Emilio, & Sorensen, Robert. "International Trade, International Investment and Industrial Profitability of U. 8. Manufacturing, " Southern Economie -.. .- -‘i‘ Joutget, Vol. AZ, No. 3 (Jan. 1976), PP. u25- M34. Peera, N. "Domestic Consumption and the Decline in India' s Jute Manufacture EXports, " Economia Internazionale, -u‘ .—._—---—-‘.—--_—__— Vol. 32, No. 2- 3 (May- -Aug, 1979) PP- 301- 308. Penson, John B. Jr. , & Webb, Kerry. "Gross National Product at Full Employment, " Federal Reserve Bank of Kansas City Econemte Review, Vol. 66, No. 6 (June 1981), pp. 3-150 Perloff, J. M., & Wachter, M. L. "A Production Function- Nonaccelerating Inflation Approach to Potential Output: Is Measured Potential Output Too High?" Carnegie- Rochester Conference Series on Public Poliey, Vol. 10, 1979 Perry, George L. "Potential Output and Productivity," Brookinge E§B§£§ 9g Econemic Activity, 1:1977, pp. 11- 219 Plosser, Charles I., & Schwert, G. William. "Potential GNP: Its Measurement and Significance, A Dissenting Opinion, " Carnegie- Rochester Conference Series on Public -.-__—-r-_—i_.—.— _.——. Policy, Vol.10, Supplementary series to the Journal of Monetety Eeepemle§,1979, pp. 179-186. Raddock, Richard D., & Forest, R. "New Estimates of Capacity Utilization: Manufacturing and Materials, " Federal Reserve Beltetin, Vol. 62, No. 11 (Nov. 1976), pp. 892- 905. Rasche, Robert R., & Tatom, John A. "The Effects of the New Energy Regime on Economic Capacity, Production, and Prices," Federal Reserve Bank Lf St. Legte Reviey, Vol. 59, No. 5 (May 1977), pp. 2- 12. Rasche, Robert R., & Tatom, John A. "Energy Resources and Potential GNP," Federal Reserve Bank Lf St. Louis -..—...... Revtey, Vol. 59, No. #6 (June 1977), pp. "10- 24. Rasche, Robert H., & Tatom, John A., "Potential Output and Its Growth Rate--The Dominance of Higher Energy Costs in the 19705 " U S E291991111 0.219111; 3811111111 the Utiltzetiep Gap. St. Louis: Washington University St. fofiis‘Cénter for the Study of American Business, December 1979. Renton, G. A. "Forecasting British EXports of Manufactures to Industrial Countries," National Institute Eeonomic --«Q ----— Reviey, No. 42 (Nov.1967),PP-35-51. Sato, Kazuo. "The Demand Function for Industrial EXports: A Cross-Country Analysis," Reviee e: Eeepemtee egg Stattettee, Vol. 59, No. A (Nov. 1977), pp. 456-46”. Smyth, D. J. "StOp-Go and United Kingdom EXports of Manufactures," Bulletin of the Oxford University Institute of Economics and Statistics, VoL 30, 1968, Steuer, M. D., Ball, R. J., & Eaton, J. R. "The Effect of Waiting Times on Foreign Orders," Economica, New --- Series, Vol. 33, No. 132 (Nov. 1966), pp. 387- U03. Stone, Joe A. "Price Elasticities of Demand for Imports and Exports: Industry Estimates for the U. S. the E. E. C., and Japan," Review Lf Eanomics and Statistics, Vol. 61, No. 2 (May 1979), pp. 307- 312. 7“- 220 Swamy, D. S. "A Quarterly Econometric Model of Demand for and Supply of EXports," Indiag Economte Revteg, N.S. Vol. 1 (April 1966), pp. 79-103. United Nations. Department of International Economic and Social Affairs, Statistical Office of the United Nations, Statistical Papers, Series D. Commeglty 12999 Stetistics. New York, various issues. ---—--.— United Nations. Department of International Economic and Statistics. New York, various issues. -—m—- U-S- Bureau of the CenSUS- 999991991999 99 99199999 919999 999 199999 919991119991999 U999 19 999911199 9191 9999199 19999 9999199199 1919- ashington. D c.= u.S. Government Printing Office, 1978. U.S. Bureau of the Census. Schedule B. Statistical --a‘- Exgoeted from the United States. (January 1, 1978, ..a. --- _-—--.-—- ——-——-——. edit1553.-_Washington, D.C.: USGPO, 1977. U.S. Bureau of the Census. U.S. Etports: Schedule E —._——-~ ~———.—_—- Commodity Grouptegee Sehedule E Commodity.ty gggqtty. ---—.-- _._—-. —‘—---— --. -—--—.-.-- Washington, D.C.: USGPO, various issues. U.S. Department of Commerce. Bureau of Economic Analysis. Business Statistics (1977 edition). Washington, D.C.: ----1‘ .— ——---_—- Ugdpo, 1978. U.S. Department of Commerce. Bureau of Economic Analysis. Survey e: 9955999 Beetneee. Washington, D.C.: USGPO, various issues. 91 99999919 Washington, D.C.: --‘-——. Indicators a99 99999919199 1999- USGPO:“‘various issues. U.S. Department of Commerce. Internation s. U.S. Department of Labor. Bureau of Labor Statistics. Employment and Eeggigg§. Washington, D.C.: USGPO, various issues. White, Kenneth J. "A General Computer Program for Econometric Methods--SHAZAM," Econometttee, Vol. 46, 1978. Winters, L. Alan. ."United Kingdom Exports and the Pressure of Demand: A Note," Econemic 999299;: Vol. 8” (Sep. 197A), pp. 623-628. 221 Winters, L. Alan. An Econometric Model of the Exeert Sector: U-K- vi.s_1r_>19 9929999‘566‘T9915'95165sjggc1979—“- Cambridge Studies in Applied Economics, No. 4, New York: Cambridge University Press, 1981. Yeager, Leland BI Internattenet Meeeteey Relations: Theory .-fi- .._._-_ Histetyt egg Polteyfi(8econd Edition)f‘_New YoFk: ..d- Harper and Row, 1976. HICHIGRN smIE UNIV. LIBRARIES 1|HI1111111111111111111111111111111 3129301 1030099