Economic Implications of a Potential Free Trade Agreement between India and the United States

This paper explores the economic implications of a potential free trade agreement between India and the United States. A series of simulations is conducted assuming 100 percent ad valorem equivalent tariff cuts for goods and 50 percent cuts for services. The overall impacts are likely to be positive for the United States and India. While gains from trade creation are offset by trade diversion on the import side, both countries appear to gain from improved access on the export side. The United States is likely to gain largely through terms of trade improvements for its goods and services, as initial protection in India is particularly high. India would experience an expansion of exports and output, especially in textiles and apparel. As the United States and India are negotiating other free trade agreements, such as the Trans-Pacific Partnership and India's agreement with the Association of Southeast Asian Nations, the paper also explores how the effects of an India-United States free trade agreement are affected by prior free trade agreements. Adding an India-United States free trade agreement to prior agreements tends to bring additional welfare benefits to both countries. India would also gain substantially if it concluded a free trade agreement with the United States and then extended it to other partners. The results suggest that an India-United States free trade agreement might become a building block toward more liberal trade regimes.

This paper explores the economic implications of a potential free trade agreement between India and the United States. A series of simulations is conducted assuming 100 percent ad valorem equivalent tariff cuts for goods and 50 percent cuts for services. The overall impacts are likely to be positive for the United States and India. While gains from trade creation are offset by trade diversion on the import side, both countries appear to gain from improved access on the export side. The United States is likely to gain largely through terms of trade improvements for its goods and services, as initial protection in India is particularly high. India would experience an expansion of exports and output, especially in textiles and apparel. As the United States and India are negotiating other free trade agreements, such as the Trans-Pacific Partnership and India's agreement with the Association of Southeast Asian Nations, the paper also explores how the effects of an India-United States free trade agreement are affected by prior free trade agreements. Adding an India-United States free trade agreement to prior agreements tends to bring additional welfare benefits to both countries. India would also gain substantially if it concluded a free trade agreement with the United States and then extended it to other partners. The results suggest that an India-United States free trade agreement might become a building block toward more liberal trade regimes.

I. Introduction
A potential Free Trade Agreement (FTA) between India and the United States presents a number of opportunities and may bring economic benefits to both countries. While India has substantially liberalized its trade and investment regime since its economic reform, which started in 1991, it remains a relatively highly protected economy. For India, deeper economic ties with the United States may provide opportunities to continue its own economic reform; to benefit from advanced and efficient technologies through trade and investment; to improve access to the large U.S. market, particularly for India's labor-intensive goods; and to create defenses against the U.S.'s protectionist pressure, especially for its growing service exports. For the United States, after having concluded successfully a number of FTAs since the 1990s, negotiating one with India may be a logical path in pursuing further its economic interests. In particular, the U.S. may expect to improve its access to the fast growing Indian market, including for those parts of the service sector in which the U.S. is likely to have a comparative advantage.
However, many economists are skeptical about using FTAs as a way to advance trade reforms, arguing that regional or bilateral trade agreements result in discriminatory liberalization, which in turn is said to cause FTA member economies to suffer from trade diversion. The uncertainty about the implications of an India-US FTA is increased by the fact the United States and India are negotiating other FTAs such as the US-EU agreement 1 , the Trans-Pacific Partnership (TPP), and India's agreement with the Association of Southeast Asian Nations (ASEAN). A key feature of an FTA is that countries reduce barriers on a "reciprocal" basis, so negotiations require FTA parties to agree to reduce their own barriers while winning concessions from their trading partner(s). Thus, the economic impacts resulting from an FTA need to be 3 evaluated for both imports and exports, addressing such issues as trade creation and trade diversion consequences and gains resulting from access to partners' markets.
The objective of this paper is to provide a preliminary assessment of the potential economic impacts of an FTA between India and the United States. For this, we use the Global Trade Analysis Project (GTAP) model (Hertel 1996) (Version 8). 2 GTAP is a relatively standard applied Computable General Equilibrium (CGE) model which is used for a variety of applications, including for studies to evaluate, ex ante, the welfare impacts of FTAs (Hertel, Hummels, Ivanic and Roman, 2007). To ensure maximum clarity and transparency, the analysis is intentionally simple and "static" in order to address key issues such as the nature and extent of trade creation and diversion from an agreement, and the sensitivity of these effects to the presence of other preferential trade agreements.
There are two major limitations to this study. Because of our focus on simplicity and transparency, the paper does not take into account "dynamic" impacts of an FTA such as the impacts of the increased foreign direct investment (FDI) inflow, positive impacts on productivity growth resulting from access to foreign knowledge, and accelerated domestic reforms (Fukase and Winters, 2003). For the same reason, we use standard trade-weighted averages of protection, rather than the more sophisticated optimal aggregation approach outlined in Laborde, Martin and Van der Mensbrugghe (2011). Since the nature of liberalization to be undertaken is not clear at this stage, the simulation scenarios are based on a uniform assumption, i.e. 100 percent Ad Valorem Equivalent (AVE) tariff cuts for goods and 50 percent cuts for services. We use the smaller reduction in barriers on services trade because many of these barriers are qualitative and difficult to distinguish from non-discriminatory liberalization. In sum, the main focus of the 4 paper is to illustrate mechanisms through which an FTA might cause welfare changes, rather than providing a precise estimate of the aggregate impacts.
Following this introduction, Section II examines the underlying patterns of trade and protection in each country. Section III illustrates theories of preferential trade liberalization.
Section IV conducts a series of simulations. We first evaluate the potential impacts of an India-US FTA on trade, output and welfare for both countries. Then we investigate how the economic implications of an India-US FTA vary depending on prior agreements. Section V presents a brief conclusion.

Figure 3b Destinations of India's Exports 2007
9 Table 1 presents the Trade Intensity Index (TII) (Drysdale and Garnaut 1982), which indicates whether the value of trade between two countries is greater or smaller than would be expected given their relative importance in world trade by the sectors used in the analysis. 3 Despite the crudity of this measure-it does not take into consideration factors such as transport cost and other characteristics-the index gives some preliminary indication as to how trade (both imports and exports) of both partners may be reduced by the presence of trade barriers. A value of TII greater than one suggests that country i exports to county j more than would be expected and vice versa. Specifically, Where Xij is country i's exports to country j Xi is i's total exports Mj is j's total imports Mi is i's total imports, and Mw is total world imports Mi is subtracted from Mw because a country cannot export goods to itself.  India are smaller than would be expected given India's share in world trade, while India's exports to the United States are greater than would be expected. Within US exports, the trade intensities are particularly low for agricultural goods (0.5 for agriculture and 0.3 for processed agriculture) and for the products of the mining sector (0.1). Exports of transport equipment and services from the US to India are disproportionately large, with trade intensities of 2.8 4 and 1.3, respectively. Within India's exports to the U.S., trade intensities are particularly low for mining sector products (0.0), and particularly high for services (2.0).     table 2a show the barriers which apply to other U.S. trading partners-these exports to the U.S. are likely to be reduced by an India-US FTA and the associated U.S. tariff revenues to be reduced, with consequent reductions in economic welfare.
An important feature of the GTAP 8 database is the use of protection data which include a comprehensive treatment of trade preferences resulting from preferential trade arrangements (PTAs) as well as the conversion of specific tariffs for both agricultural and non-agricultural goods (Guimbard, Jean, Mimouni and Pichot 2012). Overall, the average AVE rate of protection in the U.S. is relatively low at 1.3 percent. The variation of tariff rates by source countries reflects both preferential schemes and the composition of U.S. imports. The U.S. AVE protection against India's goods of 2.7 percent is higher than average and the U.S. retains particularly high barriers in the textile and clothing sector (9.1 percent). In contrast, the U.S. grants duty-free access to NAFTA partners for their apparel and textile products and applies substantially lower tariffs against the SSA region (0.5 percent) which reflects particularly the preferential rates under the African Growth and Opportunity Act (AGOA). As the US FTAs with Morocco, Chile, and Central America all allowed for immediate duty-free access to textile and apparel products meeting the agreements' rules of origin (Lawrence and Chadha 2004), a similar agreement with India-or, better, one that does not include restrictive rules of origin-would be likely to give Indian producers of textile and apparel products a competitive advantage in the US market. Table 2b shows that India's AVE tariffs are generally high (13.8 percent on average), and are particularly high in the agricultural sectors against exports from all regions. India's AVE against US goods of 9.9 percent is lower than for India's other partners, while the tariffs against some of its trading partners are especially high, for instance, 23.3 percent against its ASEAN suppliers. This is particularly important because trade diversion is more likely to generate costs when the protection against imports from other partners exceeds that from the country for which import barriers are being reduced.
On the export side, the removal of barriers by a country's FTA partner is a key factor in reaping gains from reciprocal trade liberalization. The higher the initial barriers imposed by a country's trading partner, the larger the (potential) scope of improved market access resulting from an FTA, and the bigger the potential terms-of-trade gains to the exporter. Table 3a compares the AVE protection that US exporters encounter in the Indian market with the protection they face in other markets. Overall, US exports face on average a tariff of 5.3 percent across all export markets. The protection that US exporters face in the Indian market (9.9 percent) is nearly double the average. This structure of protection suggests that the U.S is likely to gain greatly from improved market access resulting from an India-US FTA. Similarly , table 3b compares the protection Indian exporters face in the US market and those in India's other export destinations. Indian exporters face on average 4.8 percent of protection globally against their exports of goods. Since US barriers against Indian exports (2.7 percent) are relatively low overall, India's terms of trade gains on the export side may be relatively small on averagealthough the potential volumes of trade are very large. On the other hand, since the U.S. is among India's leading export destinations (figure 3b) and the US tariff against Indian labor-15 intensive goods remains high, Indian market access to the U.S. is likely to remain an important issue in its negotiation.
Many economists believe that the potential gains from liberalization in the service sector may be larger than gains available from liberalization of the trade in goods (e.g., Gervais and Analysis of the impacts of services trade liberalization is severely constrained by the lack of reliable services trade and protection data. The estimates of services trade barriers vary widely in the literature, depending on methodologies, how the barriers are defined, which sectors are included and whether the barriers include non-discriminatory impediments or apply only against foreign suppliers. However, many economists generally view the services barriers in India to be high, and those in the U.S. to be low (e.g., Borchert, Gootiiz and Mattoo 2014;Gervais and Jensen 2013;Hufbauer, Schott and Wong 2010;Petri, Plummer, and Zhai 2012). 5 As the best available educated guesstimate at the time of writing, we used the tariff equivalents of services 16 barriers reported in Hufbauer, Schott and Wong (table B.2, 2010). According to these measures, the tariff equivalents of services barriers are estimated to be 6.03 percent for the U.S. and 68.06 percent for India. In future work, it would be desirable to consider ad valorem equivalents of services barriers of the type estimated by Jafari and Tarr (2014).

Section III. The Theory of Preferential Trade Liberalization
In this section, we outline the basic framework used to evaluate the effects of reciprocal liberalization between the United States and India (Fukase and Martin 2001). On the import side,  PdacPw. However, the gains to consumers are greater as they increase the quantity of Indian goods that they purchase. Following the decline in the domestic price, consumers move down the (compensated) demand curve for Indian goods, DIndia from initial quantity m to final quantity m'.
Consumers gain the value PdacPw that would formerly have been paid to the government in tariffs. In addition, consumer surplus increases by the area abc. The net gain to the U.S. in this market can be approximated by the shaded area abc. This is the welfare benefit from trade creation.
If the import distortion being liberalized is the only distortion in the economy, then the welfare impacts of liberalization can be analyzed by considering only the trade creation effects depicted above. If, however, there are distortions in other markets, the problem is one of the second best and the impacts of liberalization on trade flows through these barriers must be considered. Perhaps the best known type of second-best welfare effect when considering a preferential trade agreement is trade diversion. In the analytical framework used in this study, this potential source of loss is readily seen by examining conditions in the market for imports from non-partner countries, represented in figure 4b. Assuming that imports from non-partner countries are substitutes for imports from partner countries, the reduction in the price of imports from India shown in figure 4b leads to a reduction in the demand for goods from non-partner countries, shifting the demand curve for these goods from Dothers to D'others. This has adverse welfare consequences that can be measured by the tariff revenues collected on non-partner imports. The welfare loss to the US is also the resulting loss in tariff revenues, shown by the area defg.
Whether there is a net gain or loss to the importing country depends on the relative sizes of the two shaded areas. Clearly, the gains from trade creation will be larger, the higher the rate of protection initially applied on these trade flows, the more price responsive is the total domestic demand for these goods (particularly, the more substitutable are domestic and imported goods) and, if the size of the increase in trade is proportional to the initial trade volume, the larger the initial trade volume. Trade diversion costs are likely to be greater the higher the tariffs applied in the non-partner markets and the greater the reduction in the quantity of imports from these markets-a quantity that reflects both the size of the drop in domestic prices of goods from the partner, and the cross-price effect of the decline in the price of imports from the partner.
Terms of Trade (TOT) effects, which are defined as a change in export prices relative to import prices, are another key component of welfare changes, as an improvement in the terms of trade contributes to welfare gains. In the case of reciprocal liberalization, the TOT gains resulting from the improved access to the partner's market are likely to be central in evaluating the welfare consequence on the export side. In figure 4c, it is shown that the reduction in Indian tariffs on exports from the US shifts India's import demand curve for exports from ED to ED'. This causes the price received by US exporters to rise from P to P'. The result is an increase in the price received for these exports and an increase in the volume of exports from the U.S. to India (shaded area hijk).
For large traders such as the US and India, there are other potentially important terms-oftrade effects. 6 An increase in a country's import demand associated with liberalization may lead to a rise in its import prices, causing the country's TOT to deteriorate. Liberalization also causes a reduction in input costs and hence improved competitiveness and exports. If this increase in export supplies causes a decline in export prices in both partner and non-partner markets, there may be a terms of trade effect that needs to be incorporated into the analysis.

Figure 4c Terms of Trade impacts of improved access to a partner x x'
The net effect on the overall TOT depends primarily on the difference between the TOT loss resulting from trade expansion and the TOT gains from preferential access to the partner's market. All of the impacts of discriminatory trade liberalization outlined above need to be taken into account simultaneously in forming an overall assessment of the proposed approach. While diagrams of the type shown above aid understanding, they do not provide a practical basis for making an overall evaluation since many of the relationships depicted are interdependent. By contrast, quantitative models such as GTAP allow all of these effects to be taken into account at once.

Trade and Output Effects
In this section, we implement a series of simulations pertaining to a potential India-US FTA. We  Table 4b estimates the trade effects to be expected when the United States and India reduce their AVE protection for services by 50 percent. U.S. exports of services to India and those from India to the U.S. are expected to increase by 113 percent and 12 percent, respectively, which would lead to a rise in overall exports from the U.S. to India by 28 percent and from India to the U.S. by 5 percent. Interestingly, the liberalization of services would appear to impact the 22 pattern of trade in goods differently in the U.S. and in India. In the United States, the liberalization of services appears to lead to a slight contraction in goods exports, with resources reallocated toward now more profitable service sectors. In contrast, the rise in exports of services from the United States to India appears to contribute to a rise in India's exports of "goods". This is perhaps because imported services provide important inputs into the production of goods, and the increased availability of efficient services may also reduce transaction/transportation costs.
Several papers find that service imports may cause developing countries to increase their international competitiveness and facilitate exports (e.g., Cebula et al. 2011;Hoekman and Braga 1997). Analyzing U.S. trade data, Cebula et al. (2011) find that service imports from the U.S. have a significant and positive impact on goods exports to the U.S. in the case of low-income countries but not in the case of high-income countries.        India is not required to lower its tariff rates under the WTO as India's bound tariff rates are far higher than its applied rates; and an India-US FTA is likely to become a comprehensive, deep and symmetrical agreement while India has little obligation to reduce its protection at the WTO because of the special and differential treatment principle (Lawrence and Chadha 2004). The simulation results show that total Indian imports and exports both expand substantially by 26 percent and 29 percent respectively; and India's exports to the United States increase disproportionately relative to other countries (45 percent) since India continues to receive preferential market access in the US market. The rise in India's imports is especially large in the processed agricultural sector (233 percent) whereas India's exports tend to expand across sectors, because of the real exchange rate depreciation associated with reduction in trade barriers (Salter 1959). In value terms, India's trade expansion is about seven times larger relative to the base scenario, with India's total trade expanding by $138 billion (11 percent of initial GDP). Table 5 reports the output effects resulting from the simulation experiments reported in tables 4a through 4d. Columns 1-2 of table 5 reveal that the impacts of goods liberalization on output are most pronounced in textiles and apparel: while the output of this sector in India increases by 7 percent, that in the U.S. contracts by 0.9 percent. It appears that, resulting from goods liberalization, productive resources are allocated more efficiently and India is able to allocate additional resources to sectors in which it has a comparative advantage.
As a result of service trade liberalization (columns 3-4), the output of services contracts slightly in India due to increased competition from the United States. However, the increased efficiency of the services sectors contributes to an expansion of output across all goods sectors.
This result is broadly consistent with the view that opening services sectors to foreign providers is a channel through which services liberalization contributes to improved performance of downstream manufacturing sectors (e.g., Arnold et al. 2012 for India; Arnold, Javorcik and Mattoo 2011 for Czech Republic). For instance, examining the link between policy reforms in services and the production of manufacturing firms from 1993 to 2005, Arnold et al. (2012) show that India's reforms in services, which include banking, telecommunications, insurance and transport reforms, had positive effects on the productivity of manufacturing.
Combining goods and services liberalizations (columns 5-6), the performance of the Indian goods sector tends to be slightly better relative to goods market liberalization alone. For instance, the change in textile and apparel output increases from 7.1 percent to 7.8 percent. When India extends its goods and services liberalization to all its suppliers (last two columns of table   5), this strategy appears to accelerate resource reallocation across sectors in India. Relative to the baseline scenario, the expansion of the textiles and apparel sector more than doubles; other sectors such as light manufacturing, chemical, rubber and plastic products, petroleum and coke products also expand sizably; but the contractions of some sectors, for instance, that of processed agriculture, are more pronounced. An India-US FTA is also likely to promote resource reallocation in the United States, but its impact on the structure of U.S. production appears to be generally small.

Welfare Effects
In order to analyze the impacts of FTAs on economic welfare, we use a decomposition of the Equivalent Variation (EV) into allocative efficiency and terms of trade (TOT) components, following Huff and Hertel (2000). The allocative efficiency effects are further decomposed into trade creation, trade diversion effects and other allocative efficiency components. 8 The first eight columns of table 6 report the results of welfare decomposition using the same scenarios described above (the baseline scenario is reported in bold). The economic impacts from goods liberalization are positive for both the United States and India as they experience welfare gains of $2.3 billion and $0.2 billion respectively (column 1-2). The gains coming from the TOT component are especially large for the United States ($2.0 billion), mainly reflecting initially high protection in India. India appears to suffer from some trade diversion in goods 8 The GTAP model incorporates many pre-existing distortions in the forms of taxes and subsidies (Huff and Hertel 2000). The welfare changes in the model are attributed to the interactions between taxes (or subsidies) and equilibrium quantity changes taking place over the course of simulations. Among allocative efficiency effects, this paper focuses on quantifying trade creation and trade diversion type effects illustrated in figure 4ab. These effects are measured as the summation of the tariff revenue weighted by imports quantity changes, i.e., by aggregating the changes in tariff revenues within FTA (trade creation) and outside of the FTA (trade diversion) (Hertel et al. 2007). The welfare effect resulting from changes in the relative prices of savings and investment is not included in our welfare decomposition. and in Latin America (Estevadeordal, Freund and Ornelas 2008). For instance, Estevadeordal et al. (2008) find strong evidence that regional agreements induced a faster decline in external tariffs in Latin America and conclude that free trade areas are likely to be building blocks in the region.
Experiments in the last four columns of table 6 separate the baseline scenario into two parts, i.e., the effects from the US lowering its protection against India's exports (Column 9-10) and from India's doing so against U.S. exports (Columns 11-12). The results reveal that both countries lose if they only lower their own barriers without obtaining reciprocal commitments, as gains resulting from doing so appear to be reaped by their partners. Our finding suggests that FTA models which focus only on the import side are misleading, and highlights the importance of negotiating concessions by FTA partner countries.   Source: Authors' simulation results. Notes: The simulations are based on a hypothetical scenario, namely, reciprocal removal of 100 percent and 50 percent of AVE protections for goods and services respectively. a The differences in bilateral trade between the United States and India reflect the differences between "cost, insurance and freight" (CIF) and "free on board" (FOB).

IV.2 India-US FTA in the Presence of Other FTAs
The series of simulations in Section IV. Similarly, given an India-ASEAN agreement, the incremental gains from an India-US FTA rise 11 However, the results need to be interpreted in the context of our assumptions. These are estimates of the potential and do not take into account leakages from measures such as "sensitive" products that might later be excluded from liberalization. For instance, in the ASEAN-India FTA, India has excluded a number of agricultural products from liberalization, putting them in the "Exclusion List" (Hoda and Gulati 2014 loses from the erosion of its preferential access to the Indian market.  (table 2b). Thus, the loss from trade diversion by adding an India-US FTA would be smaller. 13 It is noted that the "incremental" benefit from unilateral liberalization tends to be smaller relative to the corresponding scenario without prior agreements. For instance, under scenario C5W, the incremental gain coming from unilateral liberalization of $8 billion (Panel E) is smaller than the corresponding scenario without prior agreements of $12 billion (the difference between column 6 and column 8 in table 6). The smaller net gain from unilateral liberalization conditional on more prior FTAs is not surprising, since the scope of the liberalization becomes smaller when India has already removed its protection against more FTA partners. 14 The result of the combined scenario is not reported in table 7b in order to conserve the space.

Trade Effects
B5T). In scenarios C1T-C5T under which India extends unilateral liberalization after forging FTA(s), some countries appear to experience increases in trade resulting from improved access to the Indian market.

Section V. Conclusion
This paper explored the economic implications of a potential FTA between India and the United States using an applied general equilibrium model. Since the nature of the liberalization to be adopted is unknown at this stage, the potential impacts of an FTA are evaluated under a hypothetical scenario, namely 100 percent and 50 percent Ad Valorem Equivalent (AVE) tariff cuts for goods and services respectively. The results reveal that the overall impacts of an India-US FTA could be positive both for the United States and India. While gains from trade creation tend to be offset by trade diversion on the import side, both countries appear to gain from improved access to each other's markets on the export side. The U.S. is likely to gain largely through terms of trade improvement for its goods and services as the initial protection in India is particularly high. India appears to experience an expansion of exports and of output especially in the textiles and apparel sectors. Moreover, the availability of more efficient "services" imported from the United States appears to have positive impacts on production and exports of "goods" in India.
As the United States and India are negotiating other FTAs such as the US-EU agreement, the Trans-Pacific Partnership (TPP), and India's agreement with the Association of Southeast Asian Nations (ASEAN), the paper explored how the economic implications of an India-US FTA vary depending on the existence of different prior FTAs. The results reveal that adding an India-US FTA to prior agreements tends to bring additional welfare benefits to both countries. In 39 particular, India would be likely to gain substantially if it were to conclude an FTA with the U.S. and other trading partners and then to extend its commitment to all its trading partners on an MFN basis. This is because MFN liberalization unwinds costly trade diversion and promotes a more efficient resource allocation towards sectors in which India has a comparative advantage.
Finally, since countries excluded from FTAs tend to lose since they are discriminated against in FTA parties' markets and their trade is diverted in favor of FTA members, both the United States and India appear to have an incentive to enlarge the scopes of their FTAs. All the above findings suggest that an India-US FTA may potentially become a building block towards a more liberal trade regime for both countries.
Finally, our simulation exercises are subject to a number of limitations and many related issues may be subjects for future research. First, since the base year of our simulations is 2007 and our scenarios are based on simplified assumptions, future simulations might use actual liberalization schedules (when available) along with updated data. Second, it is well known that trade-weighted averages of tariff rates tend to underestimate the impact of protection and that this in turn is likely to result in the underestimate of welfare changes. In order to overcome this problem, more refined aggregators of trade distortions of a type done by Laborde et al. (2011) would be needed.