Policy, Research, and Extemral Affairs WORKING PAPERS Internatlonal Conmodity Markets International Economics Department The World Bank March 1990 WPS 383 On the Relevance of World Agricultural Prices Yair Mundlak and Dona'd F. Larson Is it appropriate for market analysts to use intemational agricul- tural prices as a proxy for domestic prices when domestic prices are unavailable? A firm yes. The Policy. Research, and External Affairs Cosplex distributes PRE Working Papers to disseninate the findings of work in progrems and to encourage the exchange of ideas among Bank staff and all others interested in developennt issues. These papers carty the names of the authors, reflect only their views, and should be used and cited accordingly. The fundings, interpetations, and conclusions are the authore' own. They should not be atuibuted to the World Bank, its Board of Diectors, its manageaent, or any of its member countries. Policy, Resoarch, and Extemal Affairs Intornational Commodity Markets This paper - a product of the Intemational Commodity Markets Division, Intemational Economics Department- is part of a larger effort in PRE to model the global markets for primary commodities and to use these models for forecasting purposes as well as forpolicy analysis. Copies are available free from the World Bank, 1818 H Street NW, Washington DC 20433. Please contact Dawn Gustafson, room S7- 044, extension 33714 (30 pages with tables). In a free market, domestic prices on agricultural Mundlak and Larson examined the appropri- products could be expected to vary with world ateness of this substitution in measuring, say, the prices. But intervention is so common with agricultural supply response to price changes- agricultural products that prices vary between particularly in the long run. countries and gaps exist between world and domestic prices. They conclude that on the whole world prices are indeed relevant. The results - for 18 The Intemational Commodity Markets Division countries and 17 commodities - are surpris- is often forced to use intemational prices as a ingly robust, and invariant to both data sources proxy for domestic prices. But it is often and time/commodity pooling. claimed that world prices are irrelevant to agricultural development in countries that intervene in agricultural pricing. 're PRE Working Paper Series disseminates the findings of work under way in the Bark's Policy, Research. and Extemal Affairs Complex. An objective of the series is to get these fundings out quickly, even if presentations are less than fuOy polished. The futdings, interpretations. and conclusionsc in these papers do not necessarily represent official Bank policy. Ptoduced at the PRE Dissemination Center Foreword In the absence of domestic prices for primary commodities the International Commodity Markets Division is often forced to use international prices as a proxy for domestic prices. This project examines the appropriateness of using intenational prices, say in measuring agricultural supply response to price changes - particularly in a long-nm contexL This project is part of the Division's progam of research on commodity price behavior and supply responsiveness in agriculture. TABLE OF COUTEIrS Introduction. * *9***9** ** ***** ..... .. ... ..*9 .. .- l The Framework.3 Empirical Reoults.5 Pooled, Country Results.8 Decomposition by Sources... .o..... o. oo.oo... o.oo.......o.......... 8 Technical Digression................................. 0... ......ll Additional Results 1eoe...e.ooosooooooo*e*.oo6 Pooled Country Data.17 Di scussion......oo .. o.o.o..... e. ooo...... oo......... .........26 References eoo..*o.*.o*o.ooo*oooooo*oo.o29 ON THE RELEVANCE OUF CORLD AGRICULTUIRAL PICRS * by Yair Mundlak and Don Larson Introduction 1. Agricultural products are on the whole tradables and every country trades in some agricultural products. In the absence of intervention it is expected that domestic prices of such products will vary with world prices. However, it is vell known that agriculture is subjected to considerable intervention which creates a gap between world and domestic prices, and generates cross-country variations in agricultural prices. Therefore, it is often claimed that world prices are irrelevant for the development of agriculture in countries which intervene in the pricing of their agricultural products. * The authors are grateful to Ronald C. Duncan and D.C. Johnson for comments and suggestions on an earlier draft. -2- 2. It to true that intervention affects the relationship between domestic and world prices. Interventions in agriculture are well documented and it is rare to find a country which does not intervene (see for esample, McCalla, 1969; Johnson 1973; Bale and Lutz, 1981; Australian Bureau of Agricultural Economics, 1985; Anderson, Hayami and Honma, 1986; World Bank 1986). The discussion on intervention deals primarily with the effect of policies on domestic prices and the consequences for domestic production, consumption trade, welfare and the spill over to the world market. Such policies are costly and therefore the reasons for their implementation are discussed as well. There are basically two approaches to the reasoning of government policies. The first one considers policy to be endogenous within the economic system. Examples of this approach are: Bullock, 1989; Rausser and !reebairn, 1°74; Rau ser and St^nchouse, 1978; Shei and Thompsoor, 1977. The second one treats policies within a broad framework where political pressures play a dominant role and therefore the response of government to changes in the economy are strongly hindered by political considerations. Examples of this approach to agricultural policy are: Abler, 1987; Gardner, 1987; Miller, 1986; Binswanger and Scandizzo, 1983. 3. In this paper we ask a different question; to what extent are world prices transmitted to domestic prices? This is a crucial topic for understanding the relationships between domestic and world markets. It is of particular interest in studying the dynamics of world agriculture (Mundlak, 1989). - 3 - 4. The insulation of a country from world prices requires resources. As the gap between domestic and world prices increases, the cost of such policies increases accordingl.y and eventually becomes excessive. Hence, it leads one to believe that the gap is bounded and if this is the case then we should see a transmission of world prices to domestic prices. This is the working hypothesis to be tested in this paper. Its implication is discussed in the concluding section. The Framework 5. The simple framework draws on the (relative) law of one price where the domestic price, P, is expressed as a product of the world price, P , the &niGizal exchanSag rate, E, nd the tax policy S I ( 't), where t 4 rthe tAx rate. P = p*ES (1) This formulation assumes that the product is homogeneous in that world and domestic prices refer to the same product; marketing margins and other domestic non-tradable inputs are ignored. This is an unrealistic assumption and any interpretations should be modified accordingly. We return to this below. At present it is assumed th t the systematic components of the non- tradable inputs are confounded in E and S whereas a disturbance U is added to account for the transitory component. Rewriting (1), with lower case letters indicating logs, we have for commodity i in year t: -4- * p -p +e + s + u . (2) it it t it it Let the relative difference in prices be d p - p* and refer to z - e + a as the policy variable, then it it it W) 2 * where u - (O, a ), and B(zu) - E(p u) 0. 6. Given the stochastic identity (2'), the elasticity of domestic prices with respect to world prices depends on the relationship between z and p*, z(p*). For example, where government actions reduce fluctuations in world prices z'(p*) c 0. At this point, we introduce the linear (in parameters) ;Cr.ion for z(t*' arA re'e- to it a- L*e policy equation; * * Z- = I + Ep + v , where E(p v) - O. (3) it 0 it it 7. The empirical relationship between domestic and world prices is given by the regression coefficient of p on p*, referred to as the elasticity of domestic price with respect to world price: b = * E* 2 (4) b EZ pitpit / Pit and by the correlation coefficient of the two prices. The summation in (4) is over comodities and time and the variables are measured as deviations from their overall means. As explained below, the variables are unitless. Using (3), the expected value of b is evauated: E(b - I) - E(EE dt p / LE p. ) (5) it it Pit 8. When the policies are independent of the world prices, b will be nearly 1. On the other hand, b is smaller than 1 when the policies reduce the fluctuations in world prices and larger than 1 in the opposite case. The closer b is to 1, the more closdly domestic prices reflect, on average, world prices. The quantitative importance of world prices in the determination of domestic prices is measured by the contribution of p* to the variations in p. This is represented by the degree of fit, (R2), of the regression of p on p*. Empirical Results Data 9. The elasticity parameter b was estimated for 58 countries for the period 1968-78: the sample covered some 60 products. The number of products varied by countries. Products not produced in a country were excluded from the analysis. 10. The domestic prices are FAO prices described by FAO as: "Farm prices are in theory determined by farm gate or first point-of- sale transaction when farmers participate in their capacity as sellers of their own products. Of course, data may not always refer to the same selling points depending on the prevailing institutional set-up in the countries. Also different practices may prevail in regard to individual communiAies." 1/ 11. The domestic prices are converted from local currencies to US dollars using exchange rates (annual averages) published by the IMP. The world price is an export unit value calculated in nominal US dollars. It is a ratio of the total world value of exports for each of the commodities divided by the total world exported quantities for the corresponding comodities. 12. Note that in this study the domestic prices are expressed in US dollars, or as P/E in terms of (1). However, there may be a difference between the exchange rate used in converting prices from domestic currencies to dollars and between the "true" rate. Therefore E in (1), or e in (2) is viewed as a correction factor for the exchange rate and is unitless. With this interpretation, the policy variable z is unitless. It represents the proportional deviation of domestic prices from world prices. Also, note that the deviations of P it from their mean (p*0.) are unitless. They simply represent the log of the proportional change in prices. 1/ FAO Production Yearbook, 1987. -7 Table 1: ELAST!'ITY OF PRODUCER PRICE WITH RESPECT TO WORLD PRICES WITHIN BETHEEN COUNTRY POOLED I t it I t Arigentina 0.966 0.759 0.990 0.551 1.000 0.794 Australia 0.930 0.847 0.933 0.486 0.944 0.907 Austria 0.979 0.790 0.991 0.106 1.009 0.902 Bangladesh 0.715 0.630 0.710 0.074 0.731 0.748 Belgium-Lux. 0.972 0.824 0.981 0.290 0.997 0.929 Brazil 0.902 1.097 0.847 0.316 0.865 1.268 Burundi 0.862 0.579 0.884 0.128 0.901 0.667 Cameroon 0.890 0.865 0.867 0.147 0.894 1.030 Canada 0.999 0.797 1.018 0.316 1.033 0.8&' Chile 0.878 0.641 0.920 0.490 0.929 0.67, Colombia 0.922 0.648 0.944 0.017 0.972 0. Costa Rica 0.908 0.659 0.933 0.417 0.946 O0;. Cyprus 0.925 0.831 0.927 0.425 0.942 0.900 Denuark 1.037 0.944 1.033 0.208 1.055 1.080 Ecuador 0.987 0.719 1.012 0.244 1.036 0.816 Egypt 1.208 0.964 1.231 0.123 1.271 1.102 El Salv^dor 0.903 0.759 0.904 0.213 0.926 0.891 Finlan 0.967 0.636 0.998 -0.051 1.023 0.789 France 0.949 0.846 0.951 0.340 0.967 0.930 Germany F.R. ).989 0.751 1.001 0.183 1.024 0.902 Greece (.912 0.846 0.908 0.208 0.925 0.948 Guatemala 0,907 0.697 0.924 0.309 0.940 0.799 India 0.737 0.438 0.775 0.173 0.795 0.503 Ireland 1.022 0.809 1.033 0.150 1.050 0.934 Israel 0.972 0.798 0.995 0.397 1.010 0.877 Italy 0.909 0.690 0.942 0.296 0.960 0.754 Japan 0.942 1.142 0.883 -0.008 0.909 1.347 Kenya 1.064 0.750 1.091 0.294 1.112 0.858 Korea, Rep. 0.921 0.997 0.894 0.231 0.910 1.097 Malawi 0.888 0.488 0.923 -0.057 0.950 0.607 Malaysia 0.858 0.836 0.846 0.358 0.862 0.958 Mauritius 1.041 0.989 1.033 0.378 1.048 1.117 Mexico 0.985 0,654 1.036 0.373 1.055 0.724 Netherlands 0.985 0.C16 0.999 0.166 1.020 0.923 New Zealand 1.029 0.7;4 1.051 0.134 1.068 0.858 Norway 0.977 0.807 0.977 0.012 1.005 0.977 Pakistan 0.744 0.362 0.803 0.004 0.829 0.437 Panama 0.937 0.603 0.965 0.213 0.963 0.720 Peru 0.868 0.803 0.861 0.124 0.880 0.958 Philippines 0.804 0.577 0.825 0.238 0.844 0.675 Portugal 0.959 0.814 0.960 0.185 0.981 0.948 South Africa 0.972 0.626 1.005 0.147 1.028 0.721 Spain 0.928 0.821 0.931 0.326 0.947 0.914 Sri-Lanka 0.814 0.686 0.827 0.588 0.833 0.706 Sweden 0.930 0.581 0.966 0.021 0.989 0.772 Switzerland 1.039 1.051 1.018 0.111 1.037 1.198 Syrla 0.978 0.872 0.977 0.108 1.002 0.992 Tanzania 0.977 0.765 0.989 0.233 1.013 0.889 Thailand 0.897 . 774 0.891 0.130 0.918 0.939 Trinidad 1.015 0.876 1.020 0.364 1.035 1.026 Turkey 0.952 0.904 0.943 0.265 0.961 1.000 United Kingdom 0.95. 0.796 0.962 0.405 0.975 0.891 United States 1.00' 0.815 1.027 0.596 1.040 0.862 Uruguay 0.796 0.730 0.800 0.427 0.809 0.776 Venezuela 0.910 0.603 0.939 0.043 0.966 0.726 Yugoslavia 1.011 0.851 1.020 0.111 1.041 0.951 Zambia 0,893 0.720 0.905 0.326 0.916 0.787 Zimbabwe 0.956 0.698 0.97k 0.023 0.994 0.834 a8- Pooled. Country Results 13. As the policy variable is unitloss, it is possible to pool the data over all commodities for all years. The estimates of the elasticity b for the poolvd data ap-ar in column 1 of Table 1. The elasticity varies between 0.715 and 1.208 with a median of 0.945. The values for 35 out of 58 countries fall in the range of 0.9-1.0. The implication is that x, the elasticity of domestic prices with resx.i.. to the policy variable, has a median value of -0.055 (1 minus 0.945) which is indeed very small. 14. The conclusion is that world prices are transmitted to domestic prices. This is a qualitative finding. The quantitative aspect is related to th9 importance of ::^Ch tran;mission. It is to be noted that in all regressions the values of R2 are quite high. This indicates not only that world prices are transmitted, but that they also constitute a major component of the variation of domestic prices. Decomposition by Sources 15. The policy equation (3) assumes a uniform policy for all commodities and all years. This assumption may be too strong and should therefore be examined. This can be done by generalizing (3). This is done by first assuming that the policy varies by commodity. In this case, the assumption E(p v) = 0 made in (3) is violated. Therefore (3) is rewritten: -9- * Z. = W; + (I + w )Pip + v's (3i) it 0 u it * * where E(p v') = 0, and cov (wiPi ) 0, for all t. The error term, v', is now defined in accordance with (3i) and, therefore, it is amsumed to be orthogonal to p . The extension in (3i) allows for a commodity-specific deviation, wi. A direct way to estimate the importance of this extension is to compute the between-commodity regression. Letting 1 p = - Z p ' the commodity-price average over time, the between commodity T it regression coefficient is: * * 2 b(i) = Z p p / £ p (4i) 1.* 1. I. Eilb(i) -1] = w + A (i, (5i) where A(i)is a weighted average of wr, so that I h(i) = ZEl X, and = (P*;2 2 (6) The values obtained for b(i) appear in column 5 (bet%ien i) cf Table 1. The differences between these results and the pooled results are negligible. The median is 0.975, as compared with 0.945 for the pooled regression. Thus, roughly speaking, the average value of wi is 0.03. It can then be concluded that either w. are generally small, or else they differ in signs and therefore their weighted avera!,R is nearly zero. We return to this below. - 10 - 16. A similar analysis follows for an alternative specification which allows for systematic variations of policy overtime. In this case, z = Ne + (1 + I ) pi + V"I (3t) it 0 t it it, where E (p*v") = 0 and cov (t P*it) = 0 for all i. Defining p. = - E Pi as the commodity-average price, the between-time regression, with variables again written as deviations from their means, is: * * 2 b(t) = Ep p Izp (4t) .t .t St and E [b(t) -11 = 1 + A(t), (5t) *2 *2 where h(t) - Ew A , and X = (p ) /£(p ) ,a weighted t t t et .t average of X t 17. The results are reported in column 6 of Table 1. There is now a larger spread in country results. This may reflect the fact that the sample consists of eleven years only, and therefore the estimates of the between-time regression are less precise than those of the between-commodity. Nevertheless, for most countries the results do not differ much from the pooled regression. The median value of b(t) is .905. Consequently the average estimate of t is approximately -0.04, similar in magnitude and sign to that of wi. - 11 - 18. The policy equation can now be extended to allow for both commodity and year effects. Combining 3i and 3t: * it 0 i t it it * * * where E(p v...) = 0; cov (i, p it) cov ( VP it O0 for all iit. 19. Finally, an interaction term, witpit, can be added to zit, with the assumption that the covariance of wit with pit is zero. This will add additional terms to the expectations of the two between-regressions. However, in the present case, this addition is quantitatively unimportant, as we shall see below. Technical Digression 20. The foregoing analysis differs somewhat from more familiar forms of analyzing panel data. We therefore turn now to evaluate the results within a uniform framework. The reader who is interested only with the empirical results can skip this discussion and move directly to the next section. Let W, B(i), B(t), W(i), and W(it) be projection (symetric and idempotent) matrices that generate residuals. They can be defined in terms of their operation on an arbitrary vector x of order IT: Wx = (sit - x..), B(i)x = (xi. - x..), B(t)x = (x.t - x..), W(X = (Kit xi.), W(t)s = (xit - x.d) W(it)s - (sit - xi. - x. + x.) - 12 - The bracketed parentheses contain the typical elements of the vectors in question. The following identities can then be derived. W = Wti) + B(i) (7a) = W(t) + B(t) (7b) = W(i) + W(t) - w(it) (7c) B(i) + B(t) + W(it) (7d) 21. Let p and p* be the vectors of the two prices, then the regression coefficients obtained above can be derived from: a = p*Ap / p*Ap*. When A = W, B(i), B(t), the resulting estimators are b (pooled), b(i) (between commodity) and b(t) (between time) respecrively. Also, when A = W(i), W(t) and W(it), the coefficients can be referred to as: within commodity, w(i), within time, w(t), and within time and commodity, w(it), respectively. Let A and C be two arbitrary matrixes and define: r(A/C) = p*Ap* / p*Cp*. (8) It then follows that: b = r[B(i)/W] b(i) + r[B[t)/W] b(t) (9) + [1 - r(B(i)/WJ - r[B(t)/WI w(it) - 13 - where, in view of (8), r[B(i)/WJ and r[B(t)/WI are the ratios of the between comodity and between time variances to the total variance of p* respectively. Table 2 presents a decomposition of the sum of squares of p* by sources. As p* is world price, the sums of squares should be the same for all countries. However, the set of commodities analyzed varies somewhat between countries and therefore the numbers in the table differ accordingly. Taking Argentina as an example, r[B(i)/WI = 481.5/562.8 = .856, r[B(t)/Wl = 69.8/562.8 = .124. Thus, the between commodity variance dominates the other components. Also note that 1 - r[B(i)/Wl - r[B(t)/WI = .02, implying that there are hardly any variations left in the world prices after the time and commodity effects were eliminated. Consequently, using the values in Table 1, it is possible to approximate the pooled regression with .856b(i) + .124b(t) = .954, as compared to the actual value of .966 for the pooled regression. The difference is due to the interaction term that was neglected. It then follows that under the present framework, the expected value of the pooled regression is: E[b - 1] a r + r[B(i)/WI A(i) + r[B(t)/W] h(t) (10) that is, the deviation from 1 (perfect transmission) consists of an overall deviation (X), and a weighted average of commodity effects and time effects, and they are all relatively small. 22. Table 1 also reports the within estimators. Allowing for commodity effects yields the within commodity estimate (column 2). Their median value is about .78 which is somewhat lower than that of the pooled regression. Allowing for time effect results in the within time estimates (column 3) with - 14 - Table 2: SUM OF SQUARES OF WORLD PRICES Pooled Within Between Check t (1) (2) (3) (4) (5) (6) (7) Country (Total) (i) (t) (it) (1) (t) Argentina 62.8 81.3 493.0 11.5 481.5 69.8 0.0 Australia 511.0 75.1 446.6. 10.6 436.0 64.5 0.0 Austria 426.4 57.2 373.2 7.7 370.4 53.4 -0.3 Bangladesh 401.2 62.9 349.3 11.0 338.3 51.9 0.0 BelgiumrLux. 414.7 58.9 358.6 9.2 357.6 56.8 -0.7 Brazil 576.0 91.8 499.3 17.5 483.6 77.0 -0.3 Burundi 385.4 46.1 346.8 7.5 339.3 38.6 0.0 Cameroon 475.8 63.6 418.7 13.0 413.0 57.9 -0.8 Canada 370.0 53.1 323.6 6.6 316.9 46.4 0.0 Chile 386.1 68.8 325.1 9.8 318.7 61.1 -0.1 Colombia 538.3 82.3 469.5 13.5 456.1 68.8 0.0 Costa Rica 442.4 59.2 390.4 11.5 384.1 52.4 -0.4 Cyprus 363.2 56.0 310.7 7.8 309.8 53.0 -0.4 Denmark 297.1 48.5 249.5 7.9 250.9 48.3 -0.8 Ecuador 542.9 83.4 473.6 14.1 459.5 69.3 0.0 Egypt 369.1 75.3 304.4 10.6 293.8 64.7 0.0 El Salvador 457.3 64.7 401.3 12.5 393.2 56.5 -0.4 Finland 273.2 39.6 235.3 6.2 235.0 38.4 -0.4 France 459.2 69.2 399.2 9.9 388.7 60.1 -0.1 Germany 429.8 55.6 375.0 10.3 371.9 56.7 -2.0 Greece 507.6 78.5 439.7 10.9 427.6 68.0 -0.1 Guatemala 437.6 58.9 386.1 11.2 379.2 51.9 -0.4 India 519.8 84.0 449.0 15.1 436.6 70.9 -0.1 Ireland 296.2 35.1 268.1 5.8 268.5 28.1 -0.1 Israel 404.6 72.5 338.3 10.9 330.5 67.1 -0.7 Italy 453.3 84.5 377.8 11.0 370.0 75.6 -0.1 Japan 562.8 78.4 494.6 12.6 486.5 68.3 -0.1 Kenya 430.2 70.3 471.1 12.9 459.8 59.2 -0.1 Korea, Rep. 482.4 63.0 425.1 7.8 424.7 57.7 -0.4 Malawi 375.8 50.1 334.7 9.0 325.8 41.1 0.0 Malaysia 446.4 58.2 394.3 13.1 390.3 53.0 -0.9 Mauritius 301.9 36.0 272.1 6.3 265.9 29.8 0.0 Mexico 522.9 92.6 440.5 14.8 433.2 82.8 -0.4 Netherlands 322.3 56.1 269.1 7.5 268.7 53.6 -0.4 New Zealand 400.1 51.5 355.3 6.7 348.6 44.8 0.0 Norway 274.5 38.8 236.1 6.9 234.6 39.6 -1.3 Pakistan 411.6 75.2 346.2 11.6 337.2 65.5 -0.1 Panama 320,5 40.2 286.7 8.4 280.5 34.0 -0.2 Peru 556.7 90.0 483.0 16.9 462.6 74.3 -0.5 Philippines 445.0 65.4 388.9 13.2 377.8 57.1 -1.0 Portugal 516.6 69.9 453.9 12.2 448,0 63.2 -0.5 ,Sourh Africa 525.9 72.8 465.1 12.1 453.1 60.7 e. Spaiua 568.2 85.8 493.2 13.6 483.0 75.4 -0.O Sri-Lanka 443.6 57.3 396.0 9.7 386.3 47.6 0.0 Sweden 365.1 53.2 310.5 10.7 315.1 56.4 -1.7 Switzerland 351.6 46.5 310.8 6.4 301.3 41.0 -0.2 Syria 397.4 71.8 335.9 9.7 325.6 61.7 -0.1 Tanzania 535.0 78.2 469.3 14.5 457.3 65.8 -0.1 Thailand 404.4 58.5 357.6 11.9 348.3 47.3 -0.5 Trinidad 388.6 47.6 345.3 10.9 341.8 44.1 -0.8 Turkey 455.6 78.6 387.2 10.2, 377.0 68.4 0.0 United Kingdom 341.8 45.9 298.6 8.2 292.1 44.2 -1.1 Uruguay 383.0 65.0 326.5 8.5 318.0 56.5 0.0 United States 528.0 81.5 457.8 13.2 445.8 70.5 -0.3 Venezuela 420.8 64.3 367.0 11.3 355.0 53.9 -A.1 Yugoslavia 488.1 76.7 420.6 9.2 411.4 67.5 0.0 Zambia 400.9 47.8 363.6 7.6 359.7 37.3 01.1 Zimbabwe 426.0 54.1 378.8 8.8 372.0 47.3 -0.1 - 15 - a median value of .955 which is almost equal to that obtained for the pooled regression. However, allowing jointly for the two effects gives very low, and in some cases even negative, elasticities. The question is what is the relationship between these various estimates. The answer is given in terms of the identities in (7a) to (7d) above. For inetance, w(i) = pW(i)p* / p*W(i)p* = b + [b - (b(i)] r[B(i)/W(i)] (11) and E(w(i)] = (1 + w) - r[B(i)/W(i)J a(i) (12) To illustrate, using the values for Argentina taken from Tables 1 and 2: b = .9656, b - b(i) = -.0344 and r[B(i)/W(i)] = 481.6/81.3 = 5.92. Substituting these values in (11) results in the value reported for w(i) in Table 1. 1/ Thus, the reason that the within commodity estimator differs from the pooled estimator is largely due to the ratio of the two variances in question rather than due to the difference between b and b(i). A similar expression can be obtained for w(t) and w(it). The latter is of a particular interest because of its big variance with the other estimates. To illustrate this point, write: 1! The results are reported here with more decimal points than in Table 1. A minor discrepancy still exists due to rounding errors. - 16 - w(it) = pW(it)p* / p*W(it)p* = b + [b - b(i)l r[B(i)/W(it)J + lb - b(t)j r[B(t)/W(it)1 (13) and therefore, Ejw(it)J = (1 + n) - a(i) rlB(i)/W(it)J - A(t) r[B(t)/W(it)J (14) Although A(i) and A(t) are relatively small, the r's are large. In the case of Argentina, r[B(i)/W(it)J = 41.87 and r[B(t)/W(it)] = 6.07, b - b(i) = -.03443 and b - b(t) = .176. Using these values in (13) gives, aside from rounding errors, the value of w(it) presented in Table 1. Additional Results 23. As indicated earlier, A(i) is estimated by the difference b(i) - b. A reference to Table 1 indicates that, for most countries, this difference is rather small. By way of summary, the difference between the median values of b(i) and b is .03, which is small relative to the reference point of unit elasticity. This by itself does not imply that the individual in(i)s are small. They may be numerically large but of opposite signs. To shed light on this point the analysis has to be conducted for a smaller set of commodities as well as for individual commodities. 24. It is often stated that staple foods are more susceptible to intervention which insulates domestic markets from world prices. Also, commodities which are traded under some sort of cartel arrangements are - 17 - expected to show a larger gap in the variations of domestic and world prices. It is therefore of interest to analyze such commodities. Table 5 presents country results for individual commodities based on 11 observations: wheat, coffee and cocoa. For wheat, the median value is approximately 0.65, and only 8 out of the 58 countries had a coefficient smaller thsn 0.5. The median value for coffee is 0.68; for cocoa it is within the range of 0.84-0.93. The conclusion is that the policy elasticities for these commodities are negative, but on the whole they are modest and by and large world prices are well transmitted. Pooled Country Data 25. There is another, not independent question: to what extent does the world price used here represent the domestic country price. This is not a trivial question. Recall that the world price is the export unit value and as such it is not an average of domestic pricer.. After all, world trade constitutes only a small fraction of world production. To examine this question, the regression is estimated with all countries pooled together. The analysis is greatly simplied when the sample is balanced, in the sense that there are no missing observations. As not all countries grow the same crops every year, a subsample was selected which consists of 17 commodities, 18 countries and ll years, altogether 3366 observations. In such an analysis the individual countries serve as repeated observations, because they all face the same world price Pit, for commodity i in year t. The pooled elasticity for this sample is .976, with R2 = .729. - 18 - Table 3: CEREALS ONLY: ELASTICITY OF PRICE WITH RESPECT TO WORLD PRICES WITHIN BETWEEN COUNTRY POOLED I t it i t Argentina 0.870 0.911 0.859 0.160 0.865 0.967 Australia 0.975 0.975 0.973 0.743 0.975 0.992 Austria 0.916 0.742 0.926 -0.045 0.932 0.799 Bangladesh 0.685 0.876 0.245 -0.084 0.268 0.908 Belgium-Lux. 0.696 0.714 0.678 0.714 0.694 0.887 Brazil 0.786 0.901 0.426 0.231 0.453 0.960 Burundi 0.682 0.686 0.684 0.830 0.675 0.680 Cameroon 0.909 1.142 0.034 0.031 0.282 1.474 Canada 0.919 0.988 0.907 -0.018 0.913 1.067 Chile 0.650 0.949 0.537 2.103 -0.336 0.704 Colombia 0.558 0.672 0.302 0.508 0.286 0.677 Costa Rica 0.572 0.596 0.543 0.646 0.509 0.592 Cyprus 0.645 0.657 0.296 0.280 0.653 0.864 Denmark 0.849 0.980 0.201 0.222 0.202 1.131 Ecuador 0.613 0.774 0.172 0.132 0.175 0.796 Egypt 0.483 0.739 -0.083 0.109 -0.098 0.761 El Salvador 0.642 0.613 0.548 0.074 0.711 0.738 Finland 0.596 0.592 0.318 0.017 0.621 0.718 France 0.812 0.751 0.814 0.108 0.820 0.798 Germany, F.R. 0.827 0.708 0.829 0.251 0.839 0.813 Greece 0.791 0.840 0.779 0.088 0.785 0.896 Guatemala 0.685 0.693 0.579 0.325 0.654 0.759 India 0.489 0.608 0.190 0.283 0.206 0.645 Ireland 0.949 0.895 0.947 0.065 0.955 0.986 Israel 0.608 0.878 -0.253 0.722 -0.533 0.881 Italy 0.630 0.708 0.378 0.096 0.396 0.757 Japan 0.908 1.227 0.863 0.063 0.870 1.312 Kenya 0.805 0.834 0.644 0.129 0.729 0.895 Korea Rep. 1.681 1.035 0.634 0.349 0.640 1.133 Malawi 0.633 0.400 1.004 -0.070 1.079 0.418 Malaysia 0.757 0.682 0.774 0.380 1.008 0.771 Mauritius 0.648 0.879 0.000 -0.112 0.011 0.917 Mexico 0.668 0.773 0.394 0.238 0.380 0.792 Netherlands 0.615 0.671 0.204 0.060 0.336 0.778 New Zealand 0.031 0.733 1.051 0.043 1.058 0.788 Norway 0.604 0.638 0.176 -0.130 0.447 0.803 Pakistan 0.246 0.311 0.221 0.741 0.077 0.262 Panama 0.371 0.583 -0.123 0.181 -0.196 0.622 Peru 0.716 0.751 0.599 0.373 0.620 0.764 Philippines 0.430 0.613 -0.016 0.190 -0.131 0.625 Portugal 0.844 0.657 0.856 0.062 0.866 0.746 South Africa 0.881 G.611 0.906 0.046 0.913 0.652 Spain 0.879 0.718 0.892 0.196 0.898 0.756 Sri Lanka 0.729 0.830 0.508 -0.136 0.547 0.864 Sweden 0.460 0.517 0.238 0.021 0.307 0.625 Switzerland 0.864 1.114 0.838 0.177 0.842 1.188 Syria 0.905 0.859 0.962 0.282 1.014 0.878 Tanzania 0.755 0.819 0.525 0.216 0.602 0.887 Thailand 0.720 0.941 0.362 1.055 0.176 0.927 Trinidad 0.648 0.738 0.223 0.121 0.343 0.836 Turkey 0.923 0.809 1.115 -0.178 1.231 0.843 United Kingdom 0.879 0.844 0.877 0.422 0.880 0.897 United States 0.925 0.909 0.925 0.749 0.927 0.914 Uruguay 0.686 0.918 0.167 0.655 0.130 0.926 Venezuela 0.666 0.755 0.475 0.164 0.496 0.777 Yugoslavia 0.922 0.807 0.930 0.807 0.936 0.846 Zambia 0.872 0.680 1.126 0.680 1.271 0.721 Zimbabwe 0.803 0.592 1.114 -0.008 1.313 0.650 - 19 - Table 4: VEGETABLES ONLY: SUMMARY TABLE FOR WITHIN AND BETWEEN COEFFICIENTS WITHIN BETWEEN COUNTRY POOLED I t It i t Argentina 1.027 0.740 1.176 0.356 1.213 0.770 Australia 0.706 0.921 0.466 0.082 0.496 0.994 Austria 0.951 0.868 1.017 0.802 0.966 0.854 Bangladesh 0.551 0.468 0.573 -U.171 0.631 0.524 Belgium, Lux. 1.182 1.052 1.311 1.135 1.325 1.046 Brazil 1.319 1.294 1.270 -0.099 1.333 1.418 Burundi 0.826 0.360 1.083 0.234 1.117 0.369 Cameroon 0.988 0.906 0.993 0.114 1.033 0.977 Canada 0.657 0.697 0.587 0.193 0.619 0.742 Chile 0.977 0.539 1.350 0.277 1.434 0.561 Colombia 1.068 0.659 1.246 -0.350 1.319 0.740 Costa Rica 1.065 0.601 1.243 0.459 1.271 0.614 Cyprus 1.006 0.958 1.041 0.736 1.067 0.974 Denmark 1.223 1.136 1.292 1.092 1.308 1.139 Ecuador 1.148 0.809 1.295 0.067 1.345 0.865 Egypt 1.194 0.994 1.331 0.216 1.419 1.053 El Salvador 0.824 0.935 0.729 0.013 0.761 1.017 Finland 1.225 0.773 1.522 0.419 1.606 0.808 France 1.000 0.825 1.191 0.849 1.220 0.823 Germany, F.R. 0.840 0.944 0.769 0.610 0.773 1.048 Greece 0.964 0.844 1.035 0.234 1.098 0.890 Guatemala 1.187 0.632 1.394 -0.202 1.455 0.704 India 0.735 0.507 0.858 0.505 0.875 0.507 Ireland 0.544 0.937 0.275 0.674 0.249 0.961 Israel 0.826 0.720 0.919 0.386 0.934 0.734 Italy 0.792 0.838 0.726 0.559 0.740 0.859 Japan 1.370 1.158 1.427 -0.026 1.493 1.290 Kenya 0.553 0.475 0.566 -0.205 0.601 0.530 Korea, Rep. 0.954 1.015 0.815 -0.054 0.926 1.188 Malawi 0.433 0.309 0.462 -0.347 0.493 0.365 Malaysia 0.236 0.803 0.023 -0.490 0.038 0.918 Mauritius 1.132 1.256 1.008 -0.047 1.057 1.363 Mexico 0.898 0.661 1.005 0.031 1.099 0.762 Netherlands 1.307 1.012 1.528 0.250 1.634 1.074 New Zealand 0.207 0.544 -0.095 -0.146 -0.091 0.608 Norway 1.320 0.838 1.612 0.237 1.715 0.899 Pakistan 0.365 0.328 0.342 -0.508 0.408 0.389 Panama 0.939 0.529 1.091 0.254 1.122 0.553 Peru 0.966 0.838 0.991 -0.279 1.049 0.923 Philippines 0.930 0.676 1.054 0.681 1.071 0.676 Portugal 0.981 1.046 0.853 -0.019 0.959 1.175 South Africa 0.891 0.580 1.111 0.282 1.175 0.608 Spain 0.972 0.929 0.972 0.495 1.002 0.975 Sri Lanka 0.782 0.665 0.802 -0.182 0.847 0.741 Sweden 1.110 0.599 1.364 0.176 1.380 0.703 Switzerland 0.994 0.960 0.983 0.182 1.047 1.016 Syria 0-825 0.814 0.794 0.250 0.841 0.851 Tanzania 1.130 0.968 1.185 0.337 1.224 1.022 Thailand 0.798 0.354 0.987 -0.293 1.046 0.412 Trinidad 0.882 1.084 0.798 0.349 0.811 1.150 Turkey 0.851 0.970 0.651 0.264 0.684 1.016 United Kingdom 1.086 0.887 1.196 0.937 1.247 0.918 United States 0.596 0.832 0.375 0.270 0.373 0.909 Uruguay 0.405 0.485 0.292 -0.152 0.327 0.541 Venezuela 0.730 0.820 0.654 0.094 0.679 0.885 Yugoslavia 0.726 0.920 0.491 0.585 0.483 0.945 Zambia 0.766 0.759 0.788 1.169 0.770 0.726 Zimbabwe 0.752, 0.580 0.805 -0.132 0.848 0.644 - 20 - Table 5: ELASTICITY OF PRODUCER PRICES WITH RESPECT TO WORLD PRICES FOR SELECTED COMMODITIES Country Wheat Coffee Cocoa Argentina 0.70118 Australia 0.90514 Austria 0.58824 Bangladesh 0.65485 Belgiuw-Luxembourg 0.62648 Brazil O.E1396 0.65206 1.1923 Burundi 0.51779 0.67989 Cameroon 1.15186 0.53017 0.61456 Canada 0.95447 Chile 0.83626 Colombia 0.62013 0.61911 0.61176 Costa Rica 0.55367 0.94008 1.07543 Cyprus 0.47708 Denmark 0.89237 Ecuador 0.52851 0.62852 0.97872 Egypt 0.56167 El Salvador 0.62115 1.05404 0.92683 Finland 0.41253 France 0.58173 Germany 0.64565 Greece 0.71502 Guatemala 0.6868 0.86177 0.97473 India 0.40493 0.14221 Ireland 0.7066 Israel 0.82092 Italy 0.65518 Japan 1.11312 Kenya 0.77996 1.00593 Korea, Rep. 0.90314 Malawi 0.50035 0.42954 Malaysia 1.00846 0.83695 0.84433 Mauritius 0.69214 Mexico 0.58558 0.85832 0.83584 Netherlands 0.58419 New Zealand 0.70093 Norway 0.60052 Pakistan 0.09736 Panama 0.49686 0.42537 1.02345 Peru 0.70418 0.73226 1.04622 Philippines 0.60916 1.01799 0.75422 Portugal 0.42215 South Africa 0.45356 Spain 0.5463 0.55763 Sri Lanka 0.58647 0.80918 1.08503 Sweden 0.48239 Switzerland 0.90987 Syria 0.68743 Tanzania 0.63865 0.61596 0.49762 Thailand 0.9951 0.46107 Trinidad 0.72659 0.60366 0.70243 Turkey 0.70537 United Kingdom 0.70782 United States 0.95847 0.83133 Uruguay 1.15293 Venezuela 0.80533 0.05066 0.50441 Yugoslavia 0.62647 Zambia 1.18678 0.71453 Zimbabwe 0.62393 0.43381 - 21 - 26. A similar analysis for subsets of commodities gives the following elasticities for the pooled regressions, with R2 reported in the parentheses: cereals .839 (.86), vegetables .933 (.61), oilseeds, .963 (.70), fruits .700 (.70), beverages .729 (.73), fibers .719 (.73), tobacco .599 (.86), livestock, .980 (.71). 27. Turning to individual commodities, with the number of countries follow the R2 in the parentheses, the results are: rice .692 (.83; 45), barley .770 (87; 49), maize .820 (.86; 58), rye .810 (.89; 39), oats .796 (.88; 51), millet .853 (.89; 34), sorghum .858 (.82; 45), wheat .693 (.85; 58), rubber .518 (.82; 10), sugar .199 (.87; 54). 28. A potential problem of any emLpirical application is that the results emerging from the study reflect the idiosyncrasies of tha way in which the data are collected, estimated or reported rather than the underlying economic effects. This possibility is inherent in applied work and can never be eliminated. However, in an effort to check for results that are consistent and independent of the underlying data source, the procedure described above was repeated on a separate data set. Tables 6 through 8 report results obtained by repeating the procedure on 25 years of producer prices in the EC- 10 as published by Herlihy et. al. (1989). Producer prices for barley, butter, cattle, cheese, eggs, maize, milk, oats, pork, poultry, pototoes, rice, rye, sugarbeets, and wheat are included in the dati. Country coverage includes Denmark, France, Greece, Ireland, Italy, the Netherlands, the United Kingdom, and West Germany. In addition, Belgium and Luxembourg are treated as a single region. Pooled EC-aggregate results are reported, as well. The - 22 - Table 6: ELASTICITY OF PRODUCER PRICES WITH RESPECT TO WORLD PRICES FOR THE EURO'EAN COMMUNITY b Standard Adjusted Country (pooled) Error T-Score R-Squared Belgium/Luxembourg 0.979 0.023 42.78 0.85 Denmark 0.957 0.019 49.76 0.88 France 1.009 0.021 48.85 0.87 Germany F.R. 0.956 0.021 45.21 0.86 Greece 0.988 0.020 49.65 0.90 Ireland 0.909 0.029 30.93 0.81 Italy 0.991 0.020 49.51 0.87 Netherlands 0.935 0.022 42.96 0.85 United Kingdom 0.957 0.019 51.28 0.89 EC Average 0.967 0.007 134.16 0.86 I" R ii fi . '0 IOOO 00 0 0.00000000 _S ' 0 .0 0.0H 0Fw.. o oeoooo,ooo Table 8: POXi 1 P ARICE E ITUY ll WAH T MD W PD FR1; Fat S1EEM) aZ ITIlES Federal Re~~~~k lk~~~dtai iloEb CxhrDdity ltmoug Demirk France of Cermny Greece Irelmd Italy ltbherlaii KlzWIm Gmdty Barley 0.71435 0.97215 0.69096 0.68419 0.63153 0.87775 0.71782 0.76180 0.86865 0.76880 Butter 0.52201 1.10116 0.60682 0.73629 0.71555 0.65394 0.87974 1.07247 0.786C0 Cattle 0.95788 1.19329 0.88884 0.98713 0.82901 1.1050B 0.87287 0.91107 1.02313 0.97425 heese 0.92142 1.28692 0.84411 1.02179 0.77285 0.83674 1.03998 1.11623 0.980m0 FgpS 0.63558 0.91878 1.19786 0.78698 0.73947 0.78471 0.54785 0.76204 0.63434 0.77862 Maize 0.68343 0.85417 0.76880 Mflk 1.18823 1.7'128 1.23026 1.39492 0.66893 1.55054 1.39304 1.31213 1.17299 1.29359 Oats 0.77850 0.99771 0.71647 0.75335 0.74656 0.84547 0.77350 0.71949 0.81865 0.79441 Pigs 0.72207 0.81864 0.57090 0.75415 0.76728 0.64235 0.73252 0.73517 0.71788 1tu1try 0.95483 0.95039 0.60383 0.87973 0.41084 0.74830 0.53365 0.93656 0.91322 0.77015 Potatoes 1.07921 1.10471 0.90602 0.90102 0.75519 1.01842 0.94985 1.01588 0.92297 0.96148 Rye 0.84387 0.91158 0.69711 0.82241 0.64540 0.84759 0.77341 1.05246 0.84759 S9arbeeta 0.74074 0.88504 0.79767 0.71826 0.62443 0.78962 0.79526 0.81790 Wheat 0.66116 0.90719 0.61932 0.72264 0.53690 0.6982C 0.77443 0.67367 0.89776 0.76830 - 25 - producer prices were originally reported in the domestic currency of each country. Official exchange rates as reported in the same publication were used to convert the prices to a dollar denomination. World prices were derived by dividing world expo:t values by world export quantities as published by FAO. 29. While the country and commodity coverage available from the EC data is more limited than in the original data set, the EC data provide an interesting and robust check on the findings. EC agricultural policy is active, well-financed, and sophisticated in its execution and reporting mechanisms. Because it is well-financed, any wedge between domestic and international prices could be expected to be more long-lived than in lower- income countries. 30. The results in Tables 6 and 7 confirm the earlier results. Although the commodity coverage is different, the pooled elasticities for countries common to both samples are remarkably similar. The producer price elasticities for the EC countries in Table 1 range from 0.91 to 1.04, while the elasticities of Table 6 range from 0.91 to 1.01. The adjusted R2s range from .81 to .90. As with the earlier results, the commodity and time effects are small. Prom Table 7, the median for b(i) is .978 as compared to the pooled result of .967, yielding a commodity effect (wi) of 0.01. The absolute value of the time effect (11t) is slighly larger at -0.08. Within commodity and time effects are reported in Table 7 as well. Again the estimeted values are consistent with results from the larger data set. Table 8 provides elasticities for individual commodities. The results for wheat can - 26 - be directly compared to the results in Table 5 which were obtained from the larger data set. Again, despite different sample years, the results are fairly comparabls. DISCUSSION 31. What do the results show? By way of generalization, the deviation from unitary elasticity is, on the whole, surprisingly small; and while there appear to be some differences among commodities and commodity groupings, the results appear quite robust regardless of the manner in which the data are pooled or disaggregated. 32. The deviation from unitary elasticity is in part due to policy measures and in part due to domestic inputs which are not necessarily synchronized with world agricultural prices. 1/ This does not imply that policies generated with respect to particular products are not important in affecting the prices of these products. They certainly affect the price levels and whenever a country taxes agriculture the domestic prices will differ from world prices. However, the question which is of concern to us is not the existence of price intervention mechanisms, but rather whether or not these mechanisms move systematically with world prices. The evidence in this paper suggests that they do not. 1/ For an analysis of this subject see Mundlak, Cavallo and Domenech. - 27 - 33. This brings up the next question: how about policies which are not related to world prices? These, by definition, will not bias the coefficient and a unitary elasticity will be observed. What is then the role of world prices in this case? The empirical answer is given by the degree of fit of the model, that is, by the proportion of the total variance of domestic prices which is accounted for by world prices. The values are relatively high. 34. The implication of this result is that technical change (and other shocks of a more permanent nature) which originate in one country but which are big enough to affect world prices, eventually affect prices in all countries. The passive countries, which are the shock takers, cannot avoid them for very long because it is too costly to do so. Realizing this cost limitation to an autonomous policy, it seems more reasonable to use, from the outset, resources to implement the necessary structural adjustments; including the enhancement of technical change, if this is the source of the shock, rather than to delay the process through taxation. This is certainly a very general statement and it has to be properly interpreted when it comes to a particular policy; however, it is mentioned here in order to put possible implications of the analysis within a broader framework. 35. Finally, we consider here a large number of commodities. In general, the trade of a country concentrates only in a few commodities, while trade in the others may be totally unimportant. The prices of the non-traded commodities is determined by domestic supply and demand and therefore, on the surface, should be independent of world prices. The explanation for the observed dependence is basically an extension of factor-price equalization. - 28 - The prices of the traded commodities determine the prices of the specific agricultural resources such an land, capital and labor in the country. 36. To conclude, even though domestic policies affect prices, they cannot prevent the covariations of domestic prices with world prices in the long run and therefore do not change the developments caused by fundamentals. There is a simple reason for it. Price distortion is costly and public resources, just like private resources, are finite. - 29 - References Abler, D.C., (1987), "Logrolling on Farm Legislation in the U.S. House of Representatives", a Ph.D thesis, Department of Economics, The University of Chicago. Anderson, K., Y. Hayami, and M. Honma, (1986), "Growth of Agricultural Protection". In Kym Anderson, Yujiro Hyami, and others. Political Economy of Agricultural Protection: The Experience of East Asia, Sydney, Australia: George Allen & Unwin. Bale, M.D. and E. Lutz (1978), Trade Restrictions and International Price Instability, World Bank Staff Working Paper 303, Washington, D.C. _ (1981). "Price Distortions in Agriculture and Their Effects: An Internattonal Comparison", American Journal of Agricultural Economics, Vol. 63, 1:8-22. Binswanger, H. and P.L. Scandizzo (1983), Patterns of hgricultural Protection. Report ARU15, Washington, D.C.: World Bank, Agriculture and Rural Development Department, Operations Po'icy Staff. Bullock, D.S., (1989), "The Volatility of Government Transfers to U.S. Agriculture, A Political Pressure Group Approach", A Ph.D thesis, Department of Economics, The University of Chicago. Bureau of Agricultural Economic, Australia (1985). Agricultural Policies in the European Community: Their Origin, Nature and. Effects on Production and Trade. Policy Monograph 2. Canberra: Australian Government Publishing Service. Gardner, B. (1987). The Economics of Agricultural Policies, New York: McMillan. Herlihy, Michael, Stephen Magiera, Richard Henry, and Kenneth Baily (1989). Agricultural Statistics of the European Community, 1960-1985, Statistical Bulletin 770, USDA, Washington, D.C. McCalla, A.F. (1969). "Protectionism in International Agricultural Trade, 1850-1968". Agricultural History 43, 3 (July): 329-44. Miller, T.C., (1986), "Explaining Differences in Agricultural Price Policy Across Countries and Across Commodities Using a Model of Competition Between Pressure Croups". A Ph.D thesis, Departmer.t of Economics, The University of Chicago. Mundlak, Y. (1989), "Agricultural Growth and World Developments", In Maunder, A., and A. Valdes (eds.), Agriculture and Governments in an Interdependent World, Proceedings of the XX International Conference of Agricultural Economists, Dartmoutha, Aldershot, England. - 30 - -, Cavallo, D. and Domenech. R. "Economic Policies, Tradability and Sectoral Prices, Argentina, 1913-84, The World Bank Economics Review, (forthcoming). Phipps, T. (1985). Farm Policies and the Rate of Return on Investment in Agriculture. Occasional paper, Washington D.C.: American Enterprise Institute. Rausser, G.C. and J.W. Freebairn (1974), "Estimation of Policy Preference Functions: An Application to U.S. Beef Import Policy", Review of Economics and Statistics, Vol. 56, No. 4, pp. 437-449. and D.P. Stonehouse (1978), "Public Intervention and Producer Supply Response", American Journal of Agricultural Economics, Vol. 60, No. 5, pp. 885-890. Shei, S.Y. and Thompson, R.L. (1977), "The Impact of Trade Restrictions on Price Stability in the World Wheat Market" ,American Journal of Agricultural Economics, Vol. 59, (4), 628-638. The World Bank, (1986) World Development Report, 1986, Washington, D.C., The World Bank. PRE Working Paper Series Contact iw Athor DA o aa WPS360 Compounding Financial Repression Bertrand Renaud January 1990 L. Victorio with Rigid Urban Regulations: Lessons 31009 of the Korea Housing Market WPS361 Housing and Labor Market Stephen K. Mayo January 1990 L. Victorio Distortions in Poland: Linkages James I. Stein 31009 and Policy Implications WPS362 Urban Property Taxation: Lessons William Dillinger January 1990 L. Victorio from Brazil 31009 WPS363 Paying for Urban Services: A Study Dale Whfttington January 1990 L. Victorio of Water Vending and Willingness Donald T. Lauria 31009 to Pay for Water in Onitsha, Nigeria Xinming Mu WPS364 Financing Urban Services in Latin Gian Carlo Guarda January 1990 L. Victorio America: Spatial Distribution Issues 31009 WPS365 Cost and Bcnefits of Rent Control Stephen Malpezzi January 1990 L. Victorio in Kumasi. Ghana A. Graham Tipple 31009 Kenneth G. Willis WPS366 Inflation, Monetary Balances, and Robert Buckley January 1990 L. Victorio the Aggregate Production Function: Anupam Dokeniya 31009 The Case of Colombia WPS367 The Response of Japanese and U.S. Panos Varangis March 1990 D. Gustafson Steel Prices to Changes in the Ronald C. Duncan 33714 Yen-Dollar Exchange Rate WPS368 Enterprise Reform in Socialist Guttorm Schjelderup Economies: Lease Contracts Viewed as a Principal-Agent Problem WPS369 Cost Recovery Strategy for Dale Whfttington March 1990 V. David Rural Water Delivery in Nigeria Apia Okorafor 33736 Augustine Akore Alexander McPhail WPS370 Export Incentives, Exchange Rate Ismail Arslan Policy, and Export Growth in Turkey Sweder van Wijnbergen WPS371 Tariff Valuation Bases and Trade Refik Erzan February 1990 J. Epps Among Developing Countries ... Alexander Yeats 33710 Do Developing Countries Discriminate Against Their Own Trade? WPS372 Long-Term Outlook for the World Shahrokh Fardoust February 1990 J. Queen Economy: Issues and Projections Ashok Dhareshwar 33740 for the 1990s PRE Workign Paper Series Contact AhAuthor DA for pager WPS373 Are Better-off Households More Lawretice Haddad March 1990 J. Sweeney Unequal or Less Unequal? Ravi Kanbur 31021 WPS374 Two Sources of Bias in Standard Samuel Laird February 1990 J. Epps Partial Equilibrium Trade Models Alexander J. Yeats 33710 WPS375 Regional Disparities, Targeting, and Gaurav Datt March 1990 C. Spooner Poverty in India Martin Ravallion 30464 WPS376 The World Economy In the Colin I. Bradford, Jr. Mid-1990s: Alternative Patterns of Trade and Growth WPS377 Security for Development In a John Stremlau Post-Bipolar World WPS378 How Does the Debt Crisis Affect Patricio Arrau Investment and Growth? A Neoclassi- cal Growth Model Applied to Mexico WPS 379 Implications of Policy Games for Miguel A. Kiguel Issues of High Inflation Economies Nissan Liviatan WPS380 Techniques for Railway Lee W. Huff March 1990 S. Shive Restructuring Louis S. Thompson 33761 WPS381 Trade in Banking Services: Alan Gelb March 1990 W. Pitayatona- Issues for Multilateral Negotiations Silvia Sagari karn 37666 WPS382 The Indonesian Vegetable Oils Donald F. Larson March 1990 D. Gustafson Sector: Modeling the Impact of 33714 Policy Changes WPS383 On the Relevance of World Yair Mundlak March 1990 D. Gustafson Agricultural Prices Donald F. Larson 33714 WPS384 A Review of the Use of the Rational Christopher L. Gilbert Expectations: Hypothesis in Models of Primary Commodity Prices WPS385 The Principles of Targeting Timothy Besley March 1990 J. Sweeney Ravi Kanbur 31021 WPS386 Argentinas Labor Markets in an Era Luis A. Riveros March 1990 R. Luz of Adjustment Carlos E. Sanchez 39059 WPS387 Productivity and Externalities: Jaime de Melo March 1990 M. Ameal Models of Export-Led Growth Sherman Robinson 37947