November 1998                                          SCSB# 390

TRADE, POLICY AND COMPETITION:
FORCES SHAPING AMERICAN AGRICULTURE PROCEEDINGS

Chapter 9 Measuring the Effectiveness of Nonprice Promotion of U.S. Agricultural Exports Using a Supply-side Approach Juan J. Porras, H. Alan Love, and C. Richard Shumway Introduction With the implementation of the North American Free Trade Agreement and recent conclusion of the Uruguay Round of the General Agreement on Tariffs and Trade (GATT), U.S. agricultural producers now have greater access to foreign markets. But with greater market accessibility also comes new competition. U.S. producers must now compete with foreign producers not only for these new markets, but also for the domestic market. One tool that has become prevalent in this competitive struggle is nonprice product promotion. Nonprice promotional activities have become the instrument of choice for exporting countries around the world. In contrast to direct price subsidies, nonprice promotional activities are allowed under GATT and are used by producers in Australia, member countries of the European Community, and New Zealand. In some countries like France and Germany, quasi-governmental agencies are used to conduct promotional activities and counsel firms about exporting. Financing for these promotional activities is provided by producer assessments in the form of checkoff funds, user fees, and government funds. In other countries, like Spain, nonprice promotional activities are financed solely by national and regional governments. The United States also offers a number of export subsidy programs. Between 1993 and 1995, the federal government spent $9.7 billion on export promotion (U.S. Congress 1995). The U.S. Department of Agriculture maintains a number of export promotion programs and accounts for a substantial share (61 percent) of total federal government spending on export promotion (U.S. Congress 1995). U.S. Department of Agriculture programs include the Export Enhancement Program (EEP), Market Promotion Program (MPP), and the Foreign Market Development program (FMD). With the exception of the EEP, these programs are targeted at promoting sales of high-value (mainly processed) food products abroad by subsidizing advertising, store promotions, trade servicing, technical assistance, and other nonprice market development activities. U.S. agricultural export promotion subsidies may be used for branded as well as generic products. Recently, nonprice promotion programs have come under increasing political scrutiny in the United States, especially the MPP which promotes high-value agricultural products and manufactured food products abroad. In a recent Associated Press release Robert Green reports: Congressional critics have repeatedly accused the program [MPP] of underwriting the advertising budgets of companies that had bucks to spare (Green 1995). The critics' principle concern is that MPP funds have been passed from product promotion groups to companies to use for brand-specific advertising. For example, Green reports: The U.S.A. Poultry and Egg Export Council gave money to companies like McDonald's to advertise Egg McMuffins and Chicken McNuggets in the Far East... [and t]he Pillsbury Co. has received millions to promote sweet corn sales in Japan (id.). This kind of scrutiny led Congress to enact a number of program reforms in the Omnibus Budget Reconciliation Act of 1993 that restricted MPP assistance to counter the adverse effects of subsidies, import quotas, or other unfair trade practices. The reforms require that MPP funds supplement, not supplant, any private sector contributions. The 1993 reforms require producer and regional trade organizations to contribute at least 10 percent of total promotional costs for generic products, and private firms to pay at least 50 percent of the cost for promotion of branded products. However, the question remains: should any government money be used to subsidize promotional activities for U.S. food products sold abroad? To date little attention has been given to evaluating the returns from export promotion programs. The few industry-specific studies focus almost exclusively on the returns of export market promotion programs to commodity producers and report widely differing results. For example, Kinnucan et al. (1995) find a marginal rate of return associated with nonprice promotional activities for U.S. cotton of $1.13, while Williams (1985) finds a marginal rate of return of $14.00 for such activities for U.S. soybean growers. In a study of the Australian beef industry, Piggott et al. (1995) report a marginal rate of return of $1.78 for advertising activities. While Kinnucan et al. and Piggott et al. rely on synthetic elasticity estimates from previous studies, each study uses a different modeling approach. Kinnucan et al. use an Armington model modified to include the effects of advertising on demand. While the Armington approach has been widely used in modeling agricultural trade (Sarris 1981; Webb et al. 1989), recent evidence indicates that underlying assumptions implicit in Armington analysis are consistently rejected across a wide range of commodities (Alston et al. 1990; Goldstein and Khan 1985; Winters 1984). Both Williams (1985) and Piggott et al. (1995) use reduced-form demand and supply models to assess the effects of export promotion on producer welfare. While these models assume perfect competition, they do not enforce the resultant zero profit condition that should be imposed in the production sector. Also, they do not include factor markets for land and labor. These deficiencies make accurate welfare analysis impossible (Peterson et al. 1994). A common feature of export market promotion studies is their focus on estimating the effects of promotional activities on foreign demand. This is a difficult task. Market promotion programs often include a diverse set of activities ranging from mass-market advertising to demonstration projects aimed at increasing use of specific agricultural commodities. Evaluating returns from an export promotion program using the traditional approach would require estimating export demand effects of numerous promotional activities for a multiplicity of commodities and importing countries. Accurately estimating the effects of a diverse set of promotion activities on export demand would be difficult, if not impossible, and would require addressing the complex issue of whether promotion simply shifts the level of demand or has more complex effects on price and income demand elasticities. Further, since most export promotion programs are not limited to specific commodities or countries, measuring program returns would require estimating a system of export demands for numerous commodities and importing countries. In this paper we take an approach that allows us to side-step the difficult demand estimation issues. Rather than focusing on estimating the effects of market promotion programs on export demand, we focus on estimating trade effects on producer welfare given various assumptions concerning post-promotion price and advertising export demand elasticities. We assess the effectiveness of the USDA's Market Promotion Program using estimated profit functions for the U.S. farm and food processing sectors that directly link producer welfare to export sales. They are used in conjunction with a synthetic export demand function for processed agricultural products. We consider a wide range of price and advertising export demand elasticities. For each price and advertising demand elasticity combination, we compute changes in farm and food processor sector profits that would result from program termination. This approach allows us to evaluate whether a particular program will be beneficial to society by assessing the likely range of post-program price and advertising export demand elasticities. To estimate trade effects on the U.S. farm and food processing sectors, we use the profit maximization (GNP function) approach pioneered by Burgess (1974) and Kohli (1978). This method has been applied extensively to national aggregate data (e.g., Kohli 1993a, 1993b, 1991, 1983, 1978; Newman, Lavy, and Devreyer 1995; Lawrence 1989; Murphy 1982) but seldom to sectoral data (Munisamy and Roe 1995). It provides a sound theoretical framework for examining export supply and import demand based on production theory. The approach treats imports as an additional input in a profit-maximizing production process which differentiates outputs between production intended for exports and production intended for domestic use. Considering the duality between production possibility sets and profit, the sector's competitive equilibrium can be characterized by the solution to a profit maximization problem subject to technological and resource constraints. Using the profit maximization approach, a complete system of input demand and output supply functions can be derived using Hotelling's lemma. The GNP function approach not only addresses many of the shortcomings inherent in other empirical trade models, it also provides theoretically consistent supply estimates for exports and domestic bound production and demand estimates for imports and domestically-produced inputs. Importantly, it facilitates direct computation of welfare measures of trade effects for both the food processing and farm sectors. Aggregating across commodities gives a parsimonious assessment of societal benefits of promotional activities. It should also provide more reliable estimates of societal program benefits than commodity models since returns that look high for one commodity may actually come at the expense of other commodities. For example, an increase in beef export demand may hurt the export demand for pork or poultry. Since the sector models treat output for domestic consumption and exports as aggregates of all commodities produced, program benefits will not be overestimated by ignoring substitution effects among agricultural commodities. If the underlying technology permits, aggregate sectoral modeling gives a more complete assessment of the total value of increases to gainers in each U.S. sector less program cost. Consequently, it results in a more accurate overall measurement of program benefits from the taxpayers' point of view. However, welfare measures from our approach are still optimistic for a reason that affects both sector and commodity models. Generic promotional activities conducted by the United States may result in the well-known free rider problem (Forker and Ward 1993). Boosting the market for U.S. exports through the use of generic advertising may in fact increase the demand for and supply of similar products from other countries. By linking the export demand directly to aggregate U.S. exports, we implicitly ignore the possibility of free riding by other competitors and provide an optimistic assessment of the returns from export promotion programs. This paper is organized into the following sections: •GNP function approach to modeling agricultural trade. •Profit-maximization approach for U.S. agriculture •Data and variable specification •Estimation methods, results, and model simulations •Conclusions References Document Prepared by: Leigh H. Stribling, lstribli@acesag.auburn.edu Alabama Agricultural Experiment Station Auburn University