Welfare impact of Central American Free Trade Agreement.

Adhikari, Deergha Raj

1. INTRODUCTION

With an objective to eliminate tariffs and trade barriers and expand regional opportunities for the workers, manufacturers, consumers, farmers, ranchers, and service providers of all the countries, the Central American-Dominican Republic-United States Free Trade Agreement (commonly called DR-CAFTA) was signed on August 5, 2004. Originally, the agreement encompassed the United States, Costa Rica, El Salvador, Guatemala, Honduras, and Nicaragua. In 2004, the Dominican Republic joined the negotiations, and the agreement was renamed DR-CAFTA. For simplification, we will call it CAFTA. This study attempts to evaluate the impact of CAFTA on U.S. welfare. In the literature, there exist two approaches to evaluating the impact of a trade policy: ex-ante approach and ex-post (revealed preference) approach. The first approach attempts to identify policy prescriptions that are welfare enhancing. But the problem with this approach is that it assumes a functional form for the measurement of welfare. Because, on one hand, if a single welfare function is assumed for the whole country, the distributional issue will be ignored as it will be difficult to rank social outcomes on the basis of potential Pareto improvement and, on the other hand, if one welfare function is assumed for each consumer and then all such functions are aggregated to represent a country as a whole, then the aggregation problem comes on the way, as consumers have diverse choices and interests.

The second approach avoids such problems by getting rid of a functional form assumption and by looking for some kind of indicator, which can be examined to infer about the welfare impact of a trade policy. A work by Ohyama (1972) derives sufficient conditions for welfare improvement using the so-called revealed preference approach, but it also suffers from same sort of distributional problems as it also assumes a representative consumer. Later Grinols and Wong (1991) extend Ohyama's results to many-consumer case and derive sufficient conditions for a trade policy to be welfare-improving with the restriction that the welfare weights of individuals be constant. Dixit and Norman's work (1980, 1986) overcomes the limitations in Grinols and Wong's model. However, their results derive the welfare-improving condition for a trade policy rather than a trade reform policy.

Later a work by Ju and Krishna (2000), builds on Dixit and Norman's and derives sufficient conditions for a trade reform to be welfare enhancing in many consumers case. However, the main problems associated with the empirical use of their model are two folds. First, it arbitrarily chooses two periods for welfare comparison. Second, it has never been tested empirically.

There are two studies by Adhikari (2003, 2006) that examine the impact of North American Free Trade Agreement (NAFTA) on the welfare of the member countries. Other empirical studies done so far on the impact of a trade policy mostly deal with the impact of the policy on certain sector of an economy rather than the whole economy, for example, on the impact of North American Free Trade Agreement on employment by U.S. International Trade Commission (1997), on agriculture by U.S. Department of Agriculture (1997), on auto and textile industries and U.S. trade balance by DeJanvry (1996), on U.S. trade and industrial structure by U.S. Department of Agriculture (1999), etc. One study by Trela and Whalley (1994), although evaluates the welfare gains to U.S. of liberalizing trade, but focuses only on the textile trade. Thus none of these studies address the issue of welfare impact of Central American Free Trade Agreement. As such, our study aims to fill the gap in international trade literature by evaluating the welfare impact of CAFTA on U.S. economy at an aggregate level.

Section-2 of this paper will lay out the model and derive the necessary and sufficient conditions for welfare improvement, section-3 will discuss about the data, section-4 will present the analyses and the results, and section-5 will summarize the findings of this study.

2. THE MODEL

Based on the work of Ohyama (1972) and Ju and Krishna (2003), we derive the necessary and sufficient conditions for U.S. welfare improvement as following.

Necessary Condition

Suppose E ([P.sup.i], [u.sup.j]) is the expenditure function of U. S. representative consumer (which shows the minimum expenditure required to achieve jth level of utility at ith level of price), C ([P.sup.i], [u.sup.j]) is the compensated demand of U. S. representative consumer (which shows the U. S. demand at ith price when utility is held constant at jth level). The superscript 0 and 1 represent the level of price or utility in the period before and after the implementation of CAFTA (Central American Free Trade Agreement) respectively. Now, suppose the representative consumer minimizes his expenditure initially, so that the following equation will hold.

E ([P.sup.0], [u.sup.0]) = [P.sup.0],C ([P.sup.0], [u.sup.0]) (1)

Where E ([P.sup.0], [u.sup.0]) and C ([P.sup.0], [u.sup.0]) are the expenditure and consumption bundle of the representative consumer respectively and [P.sup.0'] is the price vector prevailed before the implementation of CAFTA. Adding and subtracting [P.sup.0']C ([P.sup.1], [u.sup.1]) from equation (1) yields,

E ([P.sup.0], [u.sup.0]) = [P.sup.0']C ([P.sup.0], [u.sup.0]) + [P.sup.0],C ([P.sup.1], [u.sup.1]) _ [P.sup.0']C ([P.sup.1], [u.sup.1]) (2)

Equation (2) can be reorganized as following:

E ([P.sup.0], [u.sup.0]) = [P.sup.0'][C ([P.sup.0], [u.sup.0]) - C ([P.sup.1], [u.sup.1]) + [P.sup.0']C ([P.sup.1], [u.sup.1]) (2)

Since the expenditure function shows the minimum expenditure required to attain a given level of utility at a given price level, the following must be true.

[P.sup.0']C ([P.sup.1], [u.sup.1]) [greater than or equal to] E ([P.sup.1], [u.sup.1]) (4)

Where E ([P.sup.1], [u.sup.1] is the minimum expenditure required to attain a new utility level at a new price level prevailed after the implementation of CAFTA. From (3) and (4), we have

E ([P.sup.0],[u.sub.0]) [greater than or equal to] [P.sup.0'] - [C ([P.sup.0], [u.sup.0]) - C ([P.sup.1], [u.sup.1])] + E ([P.sup.1], [u.sup.1]) (5)

For E ([P.sup.1],[u.sup.1]) to be greater than E ([P.sup.0], [u.sup.0]) or equivalently, for U. S. welfare to rise after the implementation of CAFTA, the following has to be true.

[P.sup.0][C ([P.sup.0], [u.sup.0]) - C ([P.sup.1], [u.sup.1]] < 0 (6)

Inequality (6) means, for U. S. welfare to rise after CAFTA, the new (post-CAFTA) consumption bundle must not be affordable at the old (pre-CAFTA) prices. Now, suppose that X([P.sup.i], [V.sup.j]) is the U. S. domestic supply at ith price and with jth factor endowment and that the superscript 0 and 1 indicate the price and factor endowment levels before and after CAFTA respectively.

Since (6) is a necessary condition, for [P.sup.0][X ([P.sup.0], [V.sup.0]) - X ([P.sup.1], [V.sup.1])] > 0, the following is also a necessary condition:

[P.sup.0] [C ([P.sup.0], [u.sup.0]) - C ([P.sup.1], [u.sup.1])] - [P.sup.0]'[X ([P.sup.0], [V.sup.0]) - X ([P.sup.1], [V.sup.1]) < 0 (7)

The necessary condition in (7) can be rewritten as

[P.sup.0][M ([P.sup.0], [u.sup.0], [V.sup.0]) - M ([P.sup.1], [u.sup.1], [V.sup.1]) < 0 (8)

This condition means that, for U.S. welfare to increase due to CAFTA, the value of post-CAFTA U.S. import evaluated at pre-CAFTA prices must be higher than its pre-CAFTA value. After simplifying (8), the necessary condition can be rewritten as following.

[P.sup.0][M.sup.0] - [P.sup.0'][M.sup.1] < 0 (9)

where [M.sup.0] and [M.sup.1] are the Values of U.S, imports before and after CAFTA beth evaluated at pre-CAFTA price. Now suppose the U.S. implements the Central American Free Trade Agreement (CAFTA) and lowers or eliminates tariffs and other barriers on trade with the CAFTA members. For simplification we group all the CAFTA--members into group-B and all other countries into group-C. Under Armington assumptions, the necessary condition (9) now can be written as

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (10)

The subscripts A, B, and C symbolize own country, group-B countries and group-C countries respectively. The above condition can be reorganized as following:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (11)

or

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (12)

The first expression in the square bracket in (12) is the increase in the value of U.S. imports from CAFTA-member countries, the second expression is the increase in the value of U.S. imports from non-CAFTA countries, and the third expression is the increase in the value of U.S total exports (a negative sign in front of the value of imports implies the value of exports) after CAFTA was signed, all values evaluated at pre-CAFTA prices. So, the necessary condition implies that the sum of the increase in the value of imports from CAFTA countries and non-CAFTA countries must be greater than the increase in value of total U.S. exports after the formation of CAFTA when all values are evaluated at pre-CAFTA prices. Next we develop a sufficient condition for CAFTA to be welfare enhancing.

Sufficient Condition

Following the same line of reasoning as (1) through (5), the sufficient condition can be derived as following:

[P.sup.1][C ([P.sup.1], [u.sub.1]) - C ([P.sup.0], [u.sup.0])] > 0 (13)

This condition means the consumption bundle chosen by the representative consumer before CAFTA must be affordable at post-CAFTA prices. Since (13) is a sufficient condition, for [P.sup.1][X ([P.sup.1], [V.sup.1]) - X ([P.sup.0], [V.sup.0])] > 0, the following is also a sufficient condition:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (14)

The sufficient condition in (14) can be rewritten as

[P.sup.1][M ([P.sup.1], [u.sup.1], [V.sup.1]) - M ([P.sup.0], [u.sup.0], [V.sup.0]) > 0 (15)

By simplifying (15), the sufficient condition can be rewritten as following:

[P.sup.1][M.sup.1] - [P.sup.1][M.sup.0] > 0 (16)

Under Armington assumptions, the sufficient condition (16) can be written as

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (17)

The above condition can be reorganized as following:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (18)

So, the sufficient condition implies that the sum of the increases in the value of imports from CAFTA countries and non-CAFTA countries must be greater than the increase in value of total U.S. exports after the formation of CAFTA when all values are evaluated at post-CAFTA prices.

3. THE DATA

The data on U. S. import and export prices are collected from the annual series of "International Financial Statistics" published by IMF (International Monetary Fund) and the data on U. S. imports from CAFTA-member countries as well as non-CAFTA member countries and total U. S. exports are collected from different annual series of IMF's publication called "Direction of Trade'. The data on all the variables in the model are annual and cover the years 1985-2005.

4. EMPIRICAL ANALYSIS

We choose year 2005 as the post-CAFTA year and 1990 as the pre-CAFTA year to evaluate the values of U.S. imports and exports to apply the necessary and the sufficient conditions. The reason for using year 2005 as the post-CAFTA year is that, although the CAFTA was signed on August 4, 2004, it was actually implemented in 2005. The rationale for choosing year 1990 as the pre-CAFTA year is that the IMF has chosen 1990 as the base year for the computation of import and export price indices.

We evaluate the values of U.S. imports (from CAFTA-members and non-members) and the value of U.S. exports at pre-CAFTA prices by multiplying the import/export value for each year by its corresponding import/export price index for year 1990 and dividing the product by its respective year's import/export price index. Similarly, to evaluate the values of U.S. imports and exports at post-CAFTA prices we multiply the respective value for each year by its corresponding import/export price index for year 2005 and divide the product by its respective year's import/export price index. Based on the modified data we test the necessary and the sufficient conditions as following:

Necessary Condition (inequality (12))

Left Hand Side = (17,150-4,636) + (1,558,305 - 512,384) = 12,514 + 1,045,921 = 1,058,435

Right Hand Side = (845,650.1 - 393,106) = 452,544.1

Sufficient Condition (inequality (18))

Left Hand Side = (18,865 - 5,099.6) + (1,714,135 - 563,622.4) = 13,765.4 + 1,150,512.6 = 1,164,278

Right Hand Side = (904,000 - 420,230.3) = 483,769.7

Since the value on the left hand side is greater than the value on the right hand side in both of the above inequalities, the data satisfies both necessary and sufficient conditions for the CAFTA to be welfare enhancing for the U.S. Therefore, we conclude that the CAFTA has helped improve the U.S. welfare. Intuitively the above conditions means that, aider the implementation of CAFTA, the increase in the value of total imports must be greater than increase in the value of total experts either evaluated at pre-CAFTA prices or post-CAFTA prices. Meeting these conditions ensures that there will be more goods available to the representative consumer (the country as a whole) for consumption after the implementation of CAFTA.

5. SUMMARY AND CONCLUSION

Building on the work of Ju and Krishna, we develop necessary and sufficient conditions for a trade policy to be welfare improving. We apply ex-post approach to derive those conditions. The necessary condition requires that, between the years 1990 and 2005, the sum of the increases in the values of U.S. imports from countries within Central American Free Trade Agreement (CAFTA) and from countries outside CAFTA be greater than the increase in the value of U.S. exports when all values are evaluated at pre-CAFTA import/export prices. Similarly, the sufficient condition requires that, within the same period, the sum of the increases in the values of U.S. imports from CAFTA-countries and non-CAFTA countries be greater than the increase in the value of U.S. exports when all values are evaluated at post-CAFTA import/export prices. We apply those conditions on U.S. data and found that the data meets both of these conditions. Therefore, we conclude that CAFTA is welfare-enhancing for the U.S.

References

Adhikari, Deergha R. "An Intervention Analysis of a Trade Policy," International Economics, August 2006, Vol. LIX, No. 3, pp. 277-283.

Adhikari, Deergha R. "Evaluating a Trade Policy Using a Revealed Preference Approach," International Economics, November 2003, Vol. LVI, No. 4.

De Janvry, A. (1996), "NAFTA and Agriculture, An Early Assessment," Working Paper no. 807, Gianninni Foundation, University of California, Berkeley, California.

Dixit, Avinash, and Victor Norman (1980), The Theory of International Trade (Cambridge: Cambridge University Press).

Dixit, Avinash (1986), "Gains from Trade without Lump-sum Compensation," Journal of International Economics, Vol. 21, pp. 111-122.

Enders, Walter. Applied Econometric Time Series. (John Wiley and Sons: New York), 1995.

Grinols, E. L., and K. Wont (1991), "An Exact Measure of Welfare Change," Canadian Journal of Economics, Vol. 24, pp. 428-49.

Ju, Jiandong and Kala Krishna (2000), "Necessary Conditions for Welfare Improving Reforms", Economics Letters, (2000), Vol. 67, pp. 173-178.

Ju, Jiandong and Kala Krishna (2000), "Evaluating Trade Reform with Many Consumers", Canadian Journal of Economics, Vol. 33, No. 3.

Ohyama, M. (1972), "Trade and Welfare in General Equilibrium," Keio Economic Studies, 9(2), pp. 37-73.

Trela, Irene and Whalley (1994), "Trade Liberalization in Quota Restricted Items: The United States and Mexico in Textiles and Steel," in Modeling Trade Policy. J. Francois and C. Shiells, eds. Cambridge University Press.

U.S. Department of Agriculture (1999), NAFTA. WRS-99-1. Washington, D.C.

U.S. Department of Agriculture (1997), NAFTA. WRS-97-2. Washington, D.C.

U.S. International Trade Commission (1997), "Impact of the North American Free Trade Agreement on the U.S. Economy and Industries: A Three Year Review," Washington, D.C.

DEERGHA RAJ ADHIKARI

University of Louisiana at Lafayette, Lafayette