Constituency,Ideology,and Economic Interests in U.S. Congressional Voting

Abstract

Scholarly studies of U.S. legislators’ voting behavior have concluded that constituent interests exercise only limited influence, but these conclusions may result from inadequate measurement. I develop new measures of economic interests that emphasize import/export (sectoral) cleavages in addition to business/labor (factoral) cleavages and, in the process, transcend geographic boundaries. Results of logistic regression analysis suggest that the interests of economic and nongeographic constituencies, as reflected in campaign contributions, were highly significant predictors of voting in the U.S. Congress on the U.S.–Korea Free Trade Agreement and that the import/export cleavage was more salient than the business/labor cleavage. In addition, legislators’ ideological positions with respect to national security were more significant than their partisan affiliations and more significant than their positions on other dimensions of ideology.

Keywords

U.S. Congress the U.S.–Korea free trade agreement industry interest groups campaign contributions economic diplomacy

Since the U.S. and the South Korean governments launched official negotiations to establish the U.S.–Korea free trade agreement (KORUS FTA) in 2006, it has been a central issue in U.S.–Korean relations. FTAs have been actively used by policy makers to make favorable relations with partner countries as well as to increase their government competitiveness. On June 30, 2007, the KORUS FTA was signed under President George W. Bush’s trade promotion authority (TPA). Soon after, the TPA expired and the Republican president and Republican-led Congress were replaced by Democratic President Obama and a Democrat-controlled Congress. As a result, ratification of the KORUS FTA was stalled by a sharp conflict between the two main parties. In October 2010, President Obama started additional negotiations with the Korean government and finally obtained Korean concessions on several issues, in particular on automobiles. These new conditions brought a four-year-long legislative battle to an end on October 21, 2011. Its passage was crucially aided by a combination of campaign contributions from export industries and national security considerations channeled through hawkish members of Congress.

What factors indeed led U.S. legislators to vote for or against the KORUS FTA? Existing research on congressional trade policy voting has emphasized the role of political institutions (e.g., party affiliation and committee membership) and legislators’ political ideology. In contrast, the role played by constituent interests on voting behavior has garnered relatively less attention. Even though most scholars have acknowledged the theoretical connection between constituent interests and congressional voting, quantitative analyses have largely failed to show their significant relation. They have found that such interests are individually insignificant and play only a marginal role in a legislator’s trade policy vote. However, several studies speculate that the marginal role of constituent interests may be caused by the highly simplified measures of constituent interests used in empirical models (Fordham and McKeown 2003; Gartzke and Wrighton 1998).

In studies of U.S. congressional voting, constituencies are usually understood in geographic terms because the electorate is defined geographically. However, simply defining constituent interests in geographical terms suffers from several problems (Fordham and McKeown 2003). In considering nongeographically based constituent interests, a number of studies examine the impacts of campaign contributions from political action committees (PACs) on congressional voting. However, most preceding studies assume that the cleavage of constituent interests on trade policy is formed exclusively along factoral, or class-based lines (capital vs. labor). Ignoring possible sectoral interest coalitions (exporting- vs. import-competing industries) may lead to misleading results.

To fully capture the impacts of constituent interests on congressional voting, I develop a new measure for constituent interests, which includes both geographical and nongeographical interests and considers factoral as well as sectoral coalitions. With this more precise measure, I find that constituent interests are a highly significant predictor of U.S. legislators’ voting on the KORUS FTA. Moreover, sectoral coalitions are slightly more salient than factoral ones. I also test differences between the House and Senate and find that the effect of constituent interests on floor voting is more salient in the House of Representatives as usually believed. Another notable result of this study is that representatives’ national security considerations were found to be the most significant influence on voting behavior among several types of ideologies. The cumulative findings of this study imply that the KORUS FTA was the result of sectoral coalition politics, wherein exporting industries played a key role by financing political campaigns, and national security politics, wherein U.S. political leaders were more likely to support free trade with Korea as a means of strengthening alliance ties.

The remainder of this study proceeds as follows. First, I briefly describe the KORUS FTA formation process. I then review attempts to explain votes on international trade issues and discuss past studies’ main limitations in measuring constituent interests. In the research design section, I describe a new measure of constituent interests’ influence on legislators’ voting. The resulting hypotheses are then tested by logit regressions. The conclusion discusses contributions this study makes to the study of congressional voting on public policies.

The U.S.–Korea FTA Formation Process: A Two-Level Game

The negotiation and ratification process of the KORUS FTA is a typical example of a two-level game, a political model of international negotiations first developed by Putnam (1988). According to Putnam, a trade agreement proceeds along both the intranational level and international level. At the domestic level, the chief negotiator must compromise a variety of preferences from domestic actors (e.g., interest groups) and seek an agreement that is among the various possible “wins” in his state’s “win-sets.”¹ A trade agreement is finally formed when there is an overlap between the win-sets of the partner state involved in the agreement. Putnam also asserts that greater domestic constraints (i.e., a small domestic win-set) can be a bargaining advantage in international negotiations. In this sense, the completion of the KORUS FTA was more dependent on actions by the United States than those of Korea, with the former having a greater number of veto players (i.e., a smaller domestic win-set).²

To achieve certain economic and political benefits, the Bush and Roh governments signed the agreement on June 30, 2007.³ After the agreement was signed by the executives, the FTA discussions entered the realm of domestic politics, as necessary for its ratification.⁴ However, they were subsequently blocked by conflicts between legislators, mainly those supported by the expected winners (e.g., pro-business interests and the export-oriented industries) and those whose main constituents were the expected losers (pro-labor interests and the import-competing industries). Since the United States is a relatively land- and capital-abundant country, while Korea is a relatively labor-abundant one, business and agricultural groups in the United States and labor groups in Korea tend to support the FTA, and vice versa. Considering sector-specific issues, the groups expected to most suffer losses from the FTA were the agricultural industries in Korea and the auto industries in the United States. In Korea, the agricultural industries staged several massive protests, and their movements were supported by the Korean public, who were concerned that U.S. beef import could spread “mad cow disease” in Korea. In the United States, however, much attention was focused on the automobiles given the imbalance in auto trade. Due to these domestic battles around distributional effects of the FTA, the agreement could not be passed until next presidents in both countries.

In 2010, Presidents Barack Obama and Lee Myung-bak, expressed renewed commitment to the treaty. A series of additional negotiations was conducted to find a way that makes the agreement more attractive to their constituents to pave the way for its domestic ratification. Through these renegotiations, the Obama Administration obtained Korean concessions in the automobile sector. Specifically, tariff reductions for Korean automobiles were delayed for five years, and U.S. automakers were granted broader access to the Korean market. At the same time, the United States made extra concessions in the areas of beef, pharmaceuticals, and visas (Cooper et al. 2013). On October 3, 2011, ten days before a U.S. visit by President Lee, President Obama submitted the KORUS FTA to Congress for approval. On October 12, 2011, the day of Lee’s arrival, the U.S. Congress passed the agreement (along with two other FTAs with Columbia and Panama).⁵ About a month later, on November 22, the South Korean National Assembly also ratified the FTA, and the KORUS FTA finally entered into force on March 15, 2012, ending a four-year-long legislative battle on both sides.

What factors indeed led the U.S. Congress to pass the bill? At first glance, party affiliation seems to be the main factor in legislators’ decisions on the agreement. In the 111th U.S. Congress (2009–2011), Democrats dominated both chambers. Thus, it was not expected that Congress would ratify the bill. As Republicans became the House majority in the 112th Congress, there were high expectations that the FTA would be passed by the House. As anticipated, the bill was passed on October 12, 2011. If legislators’ partisanship was the only factor that led them to vote for (or against) the KORUS FTA, however, why did the Obama Administration need to engage in additional negotiations in 2010? Could Congress have ratified the FTA without the 2010 modifications?

Without the modifications made in 2010, it would have been much more difficult to obtain ratification of the KORUS FTA by the U.S. Congress because of the decreased likelihood that both executives would find win-sets. The modifications shifted the politics surrounding the agreement debate in the United States. After the conditions on the U.S. auto industry were modified, all three U.S. automakers (Ford, General Motors, and Chrysler) and the United Auto Workers (UAW), which had previously strongly opposed the FTA, finally came out in favor of it. Remarking on the UAW’s support in particular, an Obama Administration official was quoted as saying, “It has been a long time since a union supported a trade agreement,” and thus the administration hopes for a “big, broad bipartisan vote” in the U.S. Congress in 2011 (Schneider 2010). The illustrative discussions of the KORUS FTA suggest a considerable correlation between constituent interests—along with political institutions and ideology—and legislators’ voting on the agreement. In this study, I aim to determine the extent to which votes for passage in October 2011 in the House of Representatives and in the Senate can be explained by the characteristics of legislators’ constituencies, partisanship, committee membership, and ideological positions.

Determinants of Congressional Voting Behavior

Many scholars have examined the determinants of congressional voting behavior in public policy decision-making processes. A number of scholars believe that political institutions (e.g., party affiliation and committee membership) and legislators’ ideology have dominated U.S. congressional voting behavior, arguing that they act as screens through which legislators filter information about public policies.

A number of studies have examined the nature of party affiliation and support for trade legislation (Aldrich 1995; Cox and McCubbins 2005; McArthur and Marks 1988; Rohde 1994; Tosini and Tower 1987; Weller 2009). Scholars usually argue that Democrats are likely to oppose trade liberalization while Republicans are prone to support it. There are two explanations for the main parties’ different perspectives on trade liberalization. Emphasizing legislators’ ideological stances, some argue that Democrats have supported neo-Keynesian growth strategies, for which trade liberalization is relatively unimportant. In contrast, Republicans have been in favor of the neoclassical growth model, which emphasizes the welfare-enhancing impact of trade liberalization (Nelson and Silberberg 1987; Xie 2006). The class partisanship approach argues that the two main parties have appealed to different economic and social classes. Republicans tend to be more attentive to business interests and therefore more supportive of trade liberalization. Democrats are more inclined to support labor unions and hence favor protectionist policies (Gartzke and Wrighton 1998; Lohmann and O’Halloran 1994).

Hypothesis 1: Republicans are more likely to vote for the KORUS FTA than are Democrats.

When it comes to reflecting their beliefs on trade policy, legislators’ power or influence in the U.S. Congress as committee members is also an important predictor of voting behavior (Krehbiel 1996; Romer and Snyder 1994; Shepsle and Weingast 1987). Members of the committees on trade matters (e.g., the U.S. Senate Committee on Finance and the House of Representatives Committee on Ways and Means) are often seen as having a national or international orientation to trade policy and to be less responsive to the demands of narrow constituent interests. In this sense, they are expected to support trade liberalization. In contrast, members of committees dealing with labor issues are less likely to support an FTA (e.g., the U.S. Senate Committee on Health, Education, Labor and Pension and the House of Representatives Committee on Labor, Health and Human Services, Education, and Related Agencies).

Hypothesis 2: Members of committees on trade matters are more prone to vote for the agreement than are those dealing with labor issues.

The effect of a legislator’s political ideology on congressional voting behavior is a more intricate issue because political ideology can be defined in various ways (Jackson and Kingdon 1992; Nelson and Silberberg 1987; Poole and Rosenthal 2001). As a result, several types of ideology scores have been developed. In this study, I divide all types of ideology scores into three categories: (1) pure ideology scores (liberal or conservative), (2) interest group scores (pro-labor or pro-business), and (3) special issue scores (e.g., environmental and national security consideration, etc.). In contrast to previous studies examining whether a legislator’s ideological stance is significant, I attempt to investigate more specifically which type of ideological stance is more determinant of legislators’ voting on the KORUS FTA.

With regard to the first category, it is usually expected that liberals are prone to support trade interventions to increase equity, while conservatives are more likely to oppose trade interventions on efficiency grounds. The Americans for Democratic Action (ADA) annual voting records have served as the standard measure of political liberalism, while the American Conservative Union (ACU) rating measures the conservative leanings of members of Congress.

Hypothesis 3a: Conservatives are more likely to vote for the KORUS FTA than are liberals.

Second, there are several interest group scores measuring a legislator’s attachment to a certain group. For example, the American Federation of Labor–Congress of Industrial Organizations (AFL-CIO) ratings measure how closely each politician is aligned with labor interests. The Chamber of Commerce (COC) ratings tell us how closely a legislator is tied to business interests. Therefore, a legislator having a higher AFL-CIO score should be more likely to oppose trade liberalization, while a legislator’s higher COC score indicates support.

Hypothesis 3b: Legislators tied to business interests are more likely to vote for the agreement than are those aligned with labor interests.

Third, special issue scores present how much a legislator is interested in a certain issue. One of the noteworthy issues related to the KORUS FTA is U.S. national security as discussed. The American Security Council provides a national security index (NSI) measuring how consistently a legislator votes in favor of strong national defense.

Hypothesis 3c: Legislators who place more emphasis on national security are more likely to vote for the KORUS FTA.

Constituent interests are another important determinant in congressional voting behavior. Several studies argue that policy outcomes are the result of the interactions between elected officials as suppliers and constituents as demanders (Fordham and McKeown 2003; Gartzke and Wrighton 1998; Gourevitch 1986; Grossman and Helpman 1995; Schattschneider 1935). When it comes to investigating the effects of constituent interests on congressional voting, two core issues exist: how to define constituent interests and how to determine constituent preferences.

In studies of U.S. trade policies, constituencies are usually understood in geographic terms, and then constituent interests are measured by the ratio of productive factors (e.g., the ratio of capital or land to labor) or by the employment rates of certain industries in a district or a state (Anderson and Baldwin 1987; Fordham 1998; Lindsay 1990). However, several studies argue that defining constituents in solely geographic terms suffers from at least two problems (Fordham and McKeown 2003; Jackson and Kingdon 1992; Kingdon 1989). First, the mere presence of an industry or a group of voters in a district or a state does not necessarily mean that its interests are influential. Second, the actual constituency can be much broader than the geographic district because a legislator’s decisions on policy can affect interests nationally and even internationally. For example, lobbying activities from interest groups are rarely limited to geographical regions. To measure nongeographically based constituent interest, a number of studies examine the impacts of PAC campaign contributions on congressional voting (Baldwin and Magee 2000; Beaulieu and Magee 2004; Chappell 1982).

A more complex issue is how to determine the preferences of constituents on trade policy. As policy change has distributional consequences, an FTA creates economic “winner” and “loser” groups. However, the identity of the winners and losers appears to differ significantly across space and time. Two main models provide divergent predictions about which groups will support trade liberalization. Assuming that factors of production are perfectly mobile, the Heckscher–Ohlin (HO) model expects that relatively scarce factors of production lose economically from trade liberalization while relatively abundant factors gain (Mayda and Rodrik 2005; Midford 1993; Rogowski 1989). In contrast, the Ricardo–Viner (RV) model assumes the factors of production are immobile and cannot be reallocated swiftly to more efficient sectors of the economy. Therefore, individuals’ attitudes toward trade liberalization may depend on the industry in which they are employed rather than on their factor status (Busch and Reinhardt 2003; Gourevitch 1986; Hiscox 2002). Unfortunately, there is no scholarly consensus as to which model dominates the other, and both models have been empirically supported.

One of the main reasons for these inconsistent findings lies in the time frame and range of trade policy examined (Beaulieu and Magee 2004; Hiscox 2002; Ladewig 2006). Because the sectoral model assumes no factor mobility, it is more appropriate for short run analysis. In contrast, the factor model assuming perfect mobility is more applicable in the long run. Moreover, the sectoral model is likely to have more explanatory power when the range of trade policy is relatively narrow while the factoral model is dominant in the explanation for a wide range of trade policy. I assume that the sectoral model is more likely to be dominant in U.S. congressional voting on the KORUS FTA, because the agreement has a narrow range of tariff reductions with a single country, Korea, rather than broad-based tariff reductions.

Hypothesis 4a: Constituent interests formed across industries (rather than across factors) are more likely to influence the KORUS FTA formation.

The U.S. Congress consists of two chambers, the House of Representatives and the Senate. It has been believed that the influence of constituent interests on legislators’ votes is likely to be much stronger in the House than in the Senate. On the constituents’ side, the smaller and more homogeneous House district leads its members to have fewer cross-cutting cleavages on economic issues, and therefore, it may be easier to detect constituent interests in districts (rather than in states). On the legislators’ side, Representatives in the House have a short tenure (two years), and therefore they may be more sensitive to the needs and interests of their constituents for their reelection while Senators’ six-year tenures allow them to be relatively free from reelection pressures. Moreover, Senators tend to be more influential on final chamber outcomes than Representatives; therefore, partisanship and political ideology are more likely to be salient than constituent interests in the Senate (Im and Sung 2011; Jacobson 1990; Nollen and Quinn 1994; Oleszek 1984).

Hypothesis 4b: Constituent interests tend to be more salient in the House than in the Senate.

Research Design

Dependent Variable

In this study, the dependent variable is a U.S. legislator’s vote on the KORUS FTA, coded as 1 if in favor of the agreement and 0 otherwise. To test Hypothesis 4b, I examine votes in the House as well as in the Senate. Therefore, the unit of analysis is an individual Representative or Senator in the 112th U.S. Congress.

Measuring the Effects of Constituent Interests

To fully capture the effect of constituent interests on congressional voting, I develop a new measure for constituent interests considering factoral as well as sectoral constituent coalitions and taking account of both geographical and nongeographical interests. Most preceding studies have assumed that constituent interests are formed along factoral lines, ignoring possible sectoral cleavages of exporting versus import-competing industries. This assumption may be due to the more complex pattern of preferences formed along specific industry lines and the lack of detailed industry-level datasets.

Sectoral coalition operationalization

In this study, I develop a sectoral interest index considering three key components: (1) an industry’s trade orientation, to figure out whether the industry is expected to support or oppose the KORUS FTA; (2) the industry’s employment rate in a district or in a state, to measure geographical constituent interests; and (3) the industry’s campaign contribution to a legislator, to capture nongeographical interests. With regard to industry-level measures, the aggregated level of industry should be discussed first because measures can vary dramatically according to the level of aggregation (Grimwade 2000). Given the aim of this study, determining what factors led U.S. legislators to ratify the KORUS FTA, rather than more parsimonious examinations of the U.S. congressional voting for all trade policy, I select seventeen industries that are relatively sensitive with regard to establishing the KORUS FTA. The list of seventeen industries and the reasons for selecting them are presented in Appendix A.

First, the trade orientation of an industry (exporting or import-competing) indicates whether it will support or oppose the FTA. I employ 2010 industry trade data to identify an industry as either net exporting or import-competing.⁶ The measure of industry k’s trade orientation in the United States is constructed as follows:

O_{i}^{k} = \frac{E_{i}^{k} - I_{i}^{k}}{E_{i}^{k} + I_{i}^{k}},

where $O_{i}^{k}$ is the trade orientation of industry k in state i; $E_{i}^{k}$ is industry k’s export in state i to the rest of the world; and $I_{i}^{k}$ is industry k’s import in state i from the rest of the world.⁷ This measure takes on values ranging from −1 to 1. A positive value means industry k in state i is likely to be export-competing, and therefore, prone to support the KORUS FTA. In contrast, a negative value means industry k is import-competing and, thus, more likely to oppose the agreement.

To capture the geographical constituent interests, I utilize the ratio of employment rate of industry k to total employment rates of seventeen industries in district (or state) i ( $E R_{i}^{k}$ ).⁸ Consequently, the influence of industry k’s geographical interests in district i ( $G I_{i}^{k}$ ) is calculated multiplying industry k’s employment rate in district (or state) i by its trade orientation as follows:

G I_{i}^{k} = E R_{i}^{k} \times O_{i}^{k} .

To build a more complete model of the influence of economic interests on legislator voting, nongeographical interests measured by campaign contributions need to be included. The Federal Election Commission (FEC) provides data on all PAC contributions to candidates in each electoral cycle. For a more complete delineation of PAC contributions, I used the classification coding system developed by Beaulieu and Magee (2004).⁹ Instead of using the real values of campaign contribution of an industry to a legislator, I use the ratio of an industry’s campaign contribution to the total amount of campaign contributions given to a legislator to control for the size of districts (or states). The effect of industry k’s nongeographical interests on legislator m’s voting ( $n G I_{m}^{k}$ ) is calculated, by multiplying the ratio of industry k’s campaign contribution to all contributions legislator m obtains ( $P_{m}^{k}$ ) by its trade orientation in district (or state) i as in the following:

n G I_{m}^{k} = P_{m}^{k} \times O_{i}^{k} .

In both equations, all seventeen values (from seventeen industries) are summed to create a single value for a given legislator. Finally, $G I_{i}^{k}$ and $n G I_{m}^{k}$ are added as follows:

Sectoral Coalition = \sum_{k = 1}^{17} {G I_{i}^{k} + n G I_{m}^{k}} .

Factoral coalition operationalization

As the HO model assumes, constituent interests can also be formed along factoral lines. The factoral model is relatively simple. First of all, I measure the geographical interests (GI_i) by level of education because it provides a rough-and-ready picture of factor endowment on the assumption that the distribution of other types of capital is correlated with it (Fordham and McKeown 2003).¹⁰ I then measure the nongeographical interests in the factoral model with campaign contributions to a given legislator in the following:

n G I_{m} = \frac{P_{m}^{b} - P_{m}^{l}}{P_{m}^{b} + P_{m}^{l}},

where $n G I_{m}$ is the effect of nongeographical interests on legislator m’s voting on the agreement, $P_{m}^{b}$ represents campaign contributions from business PACs to legislator m, and $P_{m}^{l}$ represents those from labor PACs. This measure ranges from −1 to 1. A positive value means that legislator m acquires more contributions from business PACs, while a negative value means that he or she obtains more contributions from labor PACs. Thus, a higher number for this measure means that legislator m is expected to vote for the agreement.

Finally, a factoral coalition index is developed adding the geographical and nongeographical interests, after standardizing their variation, as follows:

Factoral Coalition = G I_{i} + n G I_{m} .

Competing Explanations: Other Independent Variables

To test Hypothesis 1, I include Party, coded 1 if a member of Congress is Democrat. For Hypothesis 2, two committee member variables are included: Com-trade is coded 1 when a legislator is a member of the committees on trade matters, while Com-labor is coded 1 when one is a member of committees related to labor issues. To test Hypothesis 3, I include all types of ideology scores: Pro-defense (NSI), Pro-labor (AFL-CO), Pro-bus (COC), Pro-lib (ADA), and Pro-con (ACU).

Demographic Characteristics: Control Variables

It is necessary to control for the impacts of other socioeconomic features and conditions of a district (or state) that are highly stressed in the previous studies. I expect that a lower income level of the median household (Income) and a higher unemployment rate (Unemp) in a district (or a state) may be correlated with a negative stance on the KORUS FTA. As it is usually expected that the KORUS FTA will increase job opportunities for Koreans in the United States, constituent pressure might be expected for members whose districts or states have a large percentage of ethnic Koreans (Korean).¹¹ Moreover, the level of unionization is expected to be negatively related to a legislator’s voting on an FTA (Union).¹²

To reduce some of the effects of the other variables, which are potentially influential but unobservable, such as logrolling and vote trading, Nollen and Quinn (1994) suggest use of a lagged endogenous variable (e.g., previous vote for the related issue). I select congressional members’ voting for the Trade Adjustment Authority (TAA) as a lagged endogenous variable. Before the ratification of the KORUS FTA, Democrats and Republicans were battling over passage of the TAA—Democrats argued that the TAA should be approved before the agreement’s ratification while Republicans insisted on the opposite order. Pre-vote is coded 1 if a legislator voted in favor of the TAA and 0 otherwise. Descriptive statistics for all the variables are presented in Table 1.

Table 1.

Descriptive Statistics.

Variable	House of representatives (N = 433)				Senate (N = 99)
Variable	M	SD	Minimum	Maximum	M	SD	Minimum	Maximum
FTA	0.64	0.48	0	1	0.83	0.38	0	1
Sector	0.69	0.68	−1.10	1.91	0.54	0.60	−1.12	1.74
Factor	−2.24e-10	1.35	−3.54	3.57	1.32e-09	1.25	−4.24	2.93
Party	0.45	0.50	0	1	0.54	0.50	0	1
Pro-defense	75.09	24.41	20	100	62.78	32.10	0	100
pro-labor	45.66	44.56	0	100	55.07	43.08	0	100
Pro-bus	58.77	34.34	0	100	58.94	37.04	0	100
Pro-lib	41.78	43.14	0	100	55.66	38.75	5	100
Pro-con	53.55	45.26	0	100	41.91	40.21	0	100
Com-trade	0.09	0.28	0	1	0.24	0.43	0	1
Com-labor	0.09	0.29	0	1	0.22	0.42	0	1
Pre-vote	0.71	0.45	0	1	0.71	0.46	0	1
(In) Unemp	1.86	0.24	0.92	2.69	2.08	0.24	1.28	2.56
(In) Income	10.85	0.25	10.10	11.58	11.11	0.15	10.84	11.45
Korean	0.58	1.59	0	19.62	0.33	0.35	0.06	1.87
EB con	179.25	181.28	0	1,322.74	326.58	344.36	0	1,379.17
IB con	2.25	4.61	0	37.5	4.10	7.02	0	43.5
EL con	42.56	57.37	0	650	23.42	39.80	0	203.53
IL con	8.30	12.30	0	68.5	3.56	9.37	0	39

FTA = free trade agreement; EB = exporting business; IB = import-competing business; EL = exporting labor; IL = import-competing labor; KORUS FTA = U.S.–Korea free trade agreement.

Empirical Results

The results of logistic analyses are presented in Table 2.¹³ In contrast to existing studies, I find that constituent interests (sector and factor) are individually significant, and that sectoral coalitions are slightly more salient than factoral coalitions in representatives’ voting on the KORUS FTA. In the House of Representatives’ voting analyses, there are 433 observations; 73–74 percent of the votes are correctly estimated on the bill through Models 1 and 2. Model 1 tests the effects of all variables. Model 2 excludes several variables that are less likely to be significant in congressional voting, to reduce noise.

Table 2.

Logistic Estimates of the U.S. Congressional Voting on the KORUS FTA.

	House				Senate
	Logistic estimates		Change in predicted probabilities		Logistic estimates		Change in predicted probabilities
	(1)	(2)	(1)	(2)	(1)	(2)	(1)	(2)
Sector	0.73***	0.71***	0.09	0.09	0.77	1.06*	0.04	0.11
Sector	(3.27)	(3.28)			(1.16)	(2.04)
Factor	0.30**	0.37***	0.08	0.09	−0.05	0.19	−0.005	0.02
Factor	(2.20)	(3.29)			(−0.13)	(0.88)
Party	1.59		0.29		5.14		0.58
Party	(1.03)				(1.31)
Pro-defense	0.05**	0.06***	0.23	0.29	−0.01	.02**	−0.00	0.003
Pro-defense	(2.56)	(9.72)			(−0.15)	(2.08)
Pro-labor	−0.04**		−0.31		−0.02		−0.00
Pro-labor	(−2.00)				(−0.25)
Pro-bus	−0.00		−0.01		0.03		0.00
Pro-bus	(−0.08)				(1.00)
Pro-lib	−0.02		−016		−0.08		−0.01
Pro-lib	(−1.03)				(−0.98)
Pro-con	−0.02		−0.19		−0.02		−0.00
Pro-con	(−0.94)				(−0.35)
Com-trade	1.98**	2.08***	0.24	0.25	0.19		0.01
Com-trade	(3.16)	(3.42)			(0.23)
Com-labor	−0.43		−0.08		−1.46*	−1.42**	−0.16	−0.20
Com-labor	(−0.86)				(−1.69)	(−2.14)
Pre-vote	0.60		0.12		0.14		0.01
Pre-vote	(1.21)				(0.07)
(In) Unemp	−0.69	−0.92	−0.03	−0.04	1.12	0.84	0.08	0.09
(In) Unemp	(−0.97)	(−1.40)			(0.78)	(0.67)
(In) Income	0.63		0.03		3.06		0.23
(In) Income	(0.83)				(0.79)
Korean	0.11		0.03		2.51		0.19
Korean	(1.06)				(1.22)
Union					−0.16	−0.05	−0.01	−0.01
Union					(−1.39)	(−0.82)
Cons	−6.22	−2.68*			−31.24	−0.97
Cons	(−0.64)	(−1.92)			(−0.73)	(−0.36)
Pseudo R²	.43	.40			.33	.19
Prob χ²	<.01	<.01			<.01	<.01

Z scores are in parentheses. KORUS FTA = U.S.–Korea free trade agreement.

Significant at 90%. **Significant at 95%. ***Significant at 99% (two-tailed test).

For substantive interpretations, I also calculate changes in the predicted probability of a “yes” vote as the independent variable changes from ½ standard deviation (SD) below the mean to ½ SD above the mean, while holding other variables at their means. This is applied to continuous variables, while for dichotomous variables changes in probabilities are reported as the explanatory variable changes from 0 to 1. In the House models, a 1 SD change of Sector (from ½ SD below mean Sector to ½ SD above the mean) increases the probability of “yes” vote by approximately 9 percent, averaged across Models 1 and 2. Factoral interests also have a significant impact on congressional voting on the agreement—a 1 SD change of Factor raises the probability of an affirmative vote by approximately 8.6 percent.¹⁴

Another significant finding is that Pro-defense shows the greatest statistical significance among all types of ideological scores. Pro-bus, Pro-lib, and Pro-con are not significant in any case, while Pro-labor is significant in the full model but loses statistical significance in the reduced model.¹⁵ These results may stem from covariation between Pro-labor (or Pro-bus) and Factor, because these two variables both capture a factoral cleavage between business and labor. A 1 SD change of Pro-defense raises the probability that a legislator votes for the agreement by approximately 21 percent. This result implies that U.S. political leaders were more likely to support free trade with Korea as a means of strengthening alliance ties with Korea.¹⁶

Com-trade was also found to be statistically significant—if a legislator is a member of the Committee on Ways and Means, the probability of voting for the agreement increases by 24 percent, averaged across Models 1 and 2. In contrast, whether a legislator is a member of committees dealing with labor issues is not a statistically significant variable.

Party is not statistically significant—this is quite a contradictory finding from preceding studies. This result is driven by high correlations between Party and other ideology variables. Without any ideological variables, Party is highly significant, indicating that a Democrat is likely to vote against the KORUS FTA. However, a likelihood-ratio test indicates that the effect of having ideological variables (at least one of five) is significant at the .01 level (LRχ² = 22.30). Jackson and Kingdon (1992) discuss the statistical bias of using the ideology scores driven from previous votes on related issues to measure the influence of a legislator’s ideology, citing overestimation of ideology and underestimation of other variables (e.g., constituent interests and party). In this study, the ideology scores may underestimate the influence of party affiliation, but they barely reduce the impact of constituent interests.

Other demographic variables do not show statistical significance. When compared with previous findings, specifically that the percentage of Hispanics is positive and significant for the NAFTA vote (Baldwin and Magee 2000; Kang and Greene 1999), it is interesting to find that the ratio of ethnic Koreans in a district is not significant in a representative’s decision on the KORUS FTA. This implies that the Korean population is not big enough to have a significant influence on U.S. trade policy.

To test Hypothesis 4b, I also examine U.S. Senate voting on the KORUS FTA. In the Senate models, there are ninety-nine observations, and the models predict 88–92 percent of the outcomes. In Model 1, the lopsided nature of the Senate vote on the KORUS FTA (eighty-three “yes” votes vs. fifteen “no” votes) indicates that party, ideology, and constituent interests may have had little or no impact on voting decisions. It is usually expected that partisanship or national interest (rather than constituents’ specific interests) are more substantial determinants in Senate votes. In the decision on the KORUS FTA, however, party affiliation does not appear to be significant in the Senate. In Model 2, the statistical significance of Pro-defense implies that national security consideration influenced Senators’ voting on the agreement. Moreover, sectoral coalition was also more salient than factoral coalition. The impacts of committee member variables in the Senate models generated the opposite results of those found in the House models. Com-labor is statistically significant, while Com-trade does not show statistical significance.

To investigate economic influences in more detail, particularly the significance of campaign contributions (nongeographical interests), four PAC groups’ contributions are included: exporting business (EB), import-competing business (IB), exporting labor (EL), and import-competing labor (IL). I also include two geographical interest variables in sectoral coalition (Employ-trade) and factoral coalition (Education). The statistical results are presented in Table 3; all types of PAC campaign contributions were statistically significant in the House but were not significant in the Senate. These results confirm again that constituent interests play a key role in representatives’ decisions on the KORUS FTA, and that industry-level cleavages among business and labor groups exist. For example, a representative who obtained contributions from export-competing (import-competing) industry PACs, regardless of whether those industries are business or labor PACs, is likely to vote for (against) the KORUS FTA. Substantively, in the House model, 1 SD change of contributions (from ½ SD below mean contributions to ½ SD above the mean) increases the probability of “yes” vote by approximately 11 percent, averaged across all four types.

Table 3.

The Effects of PAC Contributions on U.S. Congressional Voting on the KORUS FTA.

	House		Senate
	Logit estimates	Change in predicted probabilities	Logit estimates	Change in predicted probabilities
Export/business contribution	0.003***	0.10	−0.002	−0.05
Export/business contribution	(2.61)		(−1.00)
Import/business contribution	−0.14***	−0.12	0.03	0.02
Import/business contribution	(−2.83)		(0.52)
Export/labor contribution	0.01**	0.11	0.02	0.06
Export/labor contribution	(2.02)		(1.08)
Import/labor contribution	−0.05**	−0.11	−0.05	−0.03
Import/labor contribution	(−2.26)		(−0.81)
Employ-trade	0.76***	0.09	−0.28	−0.01
Employ-trade	(3.04)		(−0.70)
Education	0.05	0.09	−0.05	−0.02
Education	(2.03)		(−0.33)
Party	1.43	0.26	7.04*	0.76
Party	(0.92)		(1.75)
Pro-defense	0.05**	0.23	−0.002	−0.01
Pro-defense	(2.53)		(−0.07)
Pro-labor	−0.03*	−0.29	−0.009	−0.03
Pro-labor	(−1.81)		(−0.15)
Pro-bus	−0.01	−0.08	0.03	0.06
Pro-bus	(−0.85)		(0.70)
Pro-lib	−0.03	−0.27	−0.13	−0.50
Pro-lib	(−1.68)		(−1.57)
Pro-con	−0.02	−0.18	−0.04	−0.11
Pro-con	(−0.90)		(−0.60)
Com-trade	1.47**	0.20	0.20	0.01
Com-trade	(2.20)		(0.22)
Com-labor	−0.45	−0.09	−1.60*	−0.16
Com-labor	(−0.88)		(−1.74)
Pre-vote	0.47	0.09	0.06	0.00
Pre-vote	(0.95)		(0.03)
(In) Unemp	−0.62	−0.03	0.37	0.01
(In) Unemp	(−0.83)		(0.21)
(In) Income	−0.36	−0.02	4.25	0.04
(In) Income	(−0.37)		(0.78)
Korean	0.07	0.02	3.86	0.09
Korean	(0.60)		(1.64)
Union			−0.17	−0.00
Union			(−1.40)
Cons	3.92		−38.17
Cons	(0.34)		(−0.69)
Pseudo R²	.46		.35
Prob χ²	<.01		<.05

Z scores are in parentheses. PAC = political action committees; KORUS FTA = U.S.-Korea free trade agreement.

Significant at 90%. **Significant at 95%. ***Significant at 99% (two-tailed test).

To more specifically examine the effects of campaign contributions on voting probabilities, I conduct several counterfactual simulations. First, I use the coefficient estimates from the model and the values for each representative to predict the probability of voting in favor of the KORUS FTA. The sum of all representatives’ probabilities of voting for the KORUS FTA yields the predicted number of favorable votes. The model predicts 99.6 percent of the actual vote. Then, each representative’s probability of voting for the KORUS FTA under five counterfactuals is predicted. All other variables are held at their actual levels, but each of four groups’ (EB, IB, EL, and IL) contributions are set to zero in each simulation. In the last simulation, all PAC contributions are set to zero.

The results shown in Table 4 shed light on how important campaign contributions from the PACs are in Congressional voting on the KORUS FTA. As expected, without IL contributions, the model predicts that twenty-seven more representatives would have voted in favor of the KORUS FTA. Twelve more representatives would have voted in favor without IB contributions. In the absence of EL contributions, about twenty-three fewer representatives would have voted for the KORUS FTA. There would have been twenty-six fewer “yes” votes without EB contributions. Consequently, without campaign contributions from import-competing industries, there would have been around thirty-nine more “yes” votes, while about forty-nine fewer representatives would have voted for the KORUS FTA without contributions from export-competing industries. These results reinforce the belief that export-competing industries play a key role in congressional voting on the KORUS FTA.

Table 4.

Counterfactual Predictions of the House Vote on the KORUS FTA.

Actual KORUS FTA vote	278
Predicted votes
with all PAC contributions	277
with no contributions	262
with no labor/export PAC contributions	254
with no labor/import PAC contributions	304
with no business/export PAC contributions	251
with no business/import PAC contributions	289
Predicted effect
EL PAC contributions on number of votes	23
IL PAC contributions on number of votes	−27
EB PAC contributions on number of votes	26
IB PAC contributions on number of votes	−12

KORUS FTA = U.S.–Korea free trade agreement; PAC = political action committees; EL = exporting labor; IL = import-competing labor; EB = exporting business; IB = import-competing business.

For a robustness check on the influence of constituent interests, I test several possible interactions. Several studies argue that a vote on trade policy could be influenced by whether the legislator hails from a safe or a marginal district (Uslaner 1998; Wink, Livingston, and Garand 1996). In contrast to previous studies arguing that electoral margin has a direct impact on congressional voting, I assume that it could mediate the effect of constituent interests—constituent interests are more likely to be salient in marginal districts than in safe districts. Theoretically, the threat of electoral defeat encourages representatives to be as responsive as possible to constituent interests, while in safe districts, the lack of electoral threat allows them more room to pursue their own policy agendas. I measure electoral margin with the margin of victory by taking the difference in the vote for the incumbent and the second-place competitor in the most recent election. Similarly, constituent interests may be conditioned by the number of years served by a legislator. As legislators have served in Congress for a longer time, they are less likely to be sensitive to constituent interests given a lower perceived threat of losing the next election.

When I include four interactive terms (Margin*Sector, Margin*Factor, Term*Sector, and Term*Factor), their coefficients are not statistically significant, as shown in Table 5. Moreover, the constituent interest variables (Sector and Factor) look as they do in models excluding the interactive terms. The results suggest that constituent interests are not conditioned by a legislator’s safety in a district (or a state). In other words, legislators vote for their constituent interests regardless of how competitive their districts are and how long they serve in Congress.¹⁷

Table 5.

Logistic Estimates of U.S. Congressional Voting on the KORUS FTA with Interactions.

	House		Senate
	(1)	(2)	(1)	(2)
Sector	0.90***	0.72**	0.87	0.49
Sector	(2.85)	(2.46)	(1.05)	(0.62)
Factor	.20	0.30	−0.64	−0.81
Factor	(0.93)	(1.49)	(−0.96)	(−1.47)
Margin × Sector	−0.01		0.01
Margin × Sector	(−0.85)		(0.42)
Margin × Factor	0.003		0.01
Margin × Factor	(0.62)		(0.86)
Term × Sector		−0.001		0.08
Term × Sector		(−0.08)		(1.39)
Term × Factor		−0.001		0.05
Term × Factor		(−0.09)		(1.62)
Party	−0.35	−0.22	0.51	0.82
Party	(−0.45)	(−0.29)	(0.22)	(0.32)
Pro-defense	0.06***	0.06***	0.04	0.04
Pro-defense	(3.54)	(3.65)	(0.98)	(1.05)
Com-trade	2.03***	2.01***	−0.07	−0.08
Com-trade	(3.28)	(3.24)	(−0.09)	(−0.10)
Com-labor	−0.51	−0.49	−1.84**	−2.11**
Com-labor	(−1.03)	(−1.00)	(−2.27)	(−2.54)
(In) Unemp	−0.66	−0.87	0.88	1.73
(In) Unemp	(−0.93)	(−1.26)	(0.62)	(1.12)
(In) Income	0.39	0.40	4.44	5.32
(In) Income	(0.53)	(0.56)	(1.17)	(1.30)
Korean	0.10	0.09	3.00	2.58
Korean	(0.98)	(0.90)	(1.37)	(1.11)
Union			−0.19**	−0.18**
Union			(−1.96)	(−1.99)
Cons	−6.98	−6.92	−50.38	−62.57
Cons	(−0.79)	(−0.80)	(−1.19)	(−1.99)
Observation	433	433	99	99
Pseudo R²	<.01	<.01	<.05	<.01
Prob χ²	.41	.41	.28	.32

Z scores are in parentheses. KORUS FTA = U.S.–Korea free trade agreement.

Significant at 90%. **Significant at 95%. ***Significant at 99% (two-tailed test).

Conclusion

The results of this study make at least two significant contributions to congressional voting analyses. First, to fully capture the influence of constituent interests on legislators’ voting for trade policy, I develop a new measure for constituent interests by considering factoral (i.e., class-based) as well as sectoral coalitions, and by taking into account geographical as well as nongeographical constituent interests. In developing a new measure, I attempt to bring the data to the proper level of observation to capture variances that exist at the district and individual levels. This more precise measure, taken together with its estimated effects, improves the confidence that the marginal role of constituent interests found in existing studies is a consequence of the highly simplified measures used. I find that constituent interests played a significant role in U.S. legislators’ voting on the KORUS FTA and that it was more likely to be a determinant in the House of Representatives rather than in the Senate. Specifically, both sectoral and factoral coalitions were present in U.S. congressional voting on the KORUS FTA, and sectoral coalitions of exporting versus import-competing industries are slightly more salient than factoral coalitions of business versus labor groups. This finding implies that the KORUS FTA is to a large extent the result of domestic political games between expected winners and losers.

Rather than investigating whether legislators’ ideology is a decisive factor in congressional voting, I attempt to more specifically examine which type of ideology is crucial. These types are represented by three ideology scores (i.e., pure ideology, interest groups, and special issue scores). I find that representatives’ national security concerns have the greatest statistical significance in influencing their voting on the KORUS FTA. This finding provides another significant implication: that the KORUS FTA is to a large extent the result of representatives’ intention to strengthen an alliance important to national security.

The cumulative findings of this study indicate that the KORUS FTA was the result of domestic as well as international political considerations. Consequently, the KORUS FTA discussions offer significant empirical evidence that FTAs are the aggregated results of two-level political considerations and are not simply based on pure economic optimality calculations. Due to this entangled interest of players at the domestic as well as the international levels, an FTA, in some cases, is not an effective tool to achieve the policy goals such as forming favorable relations or enhancing government competitiveness. The findings of this study of course could be limited to the single case. However, the way to more precisely measure the impact of constituent interests makes significant contributions to the study on U.S. congressional voting on trade policy.

Footnotes

Appendix

Given the aim of this study, determining what factors led the U.S. legislators to ratify the KORUS FTA, rather than more parsimonious examinations of the U.S. congressional voting for all trade policy, I select seventeen industries that are relatively sensitive with regard to establishing the KORUS FTA. First, I choose the top ten industries in trade between the United States and Korea, as they are more likely to be influenced by the agreement.

However, pre-existing trade patterns have an imperfect ability to explain the preferences of industries. More specifically, some sectors have little pre-existing trade, not due to lack of trade complementarity but as a result of large pre-existing trade barriers. Such sectors expect large increase of benefits (or costs) after an FTA is established. For example, even though its pre-existing volume of trade with Korea was not huge, the U.S. pharmaceutical industry strongly supported the KORUS FTA. It expected large increases in exports to Korea after the FTA was established, as compared with the pre-FTA period when Korea strongly protected its home market. The U.S. International Trade Commission (USITC) usually prepares reports assessing the possible sectoral effects of future FTAs before legislators’ vote for FTAs. Based on the USITC (2007) report, I select seven more sensitive industries, based on expectations of effects comparable with the ten industries with the greatest pre-existing trade levels. To reduce potential bias in selection, it is important to make sample balance between pro- and anti-free trade industries. In this sense, three expected winners (51, 52, and 3253) and four losers (313-6, 3341, 3342, and 3359) are selected.

Author’s Note

Replication data for this article can be viewed at .

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by National Research Foundation of Korea Grant funded by the Korean Government (NRF-2014S1A3A2044898).

Notes

References

Aldrich

John

. 1995. Why Parties? The Origin and Transformation of Party Politics in America. Chicago: University of Chicago Press.

Anderson

Kym

Baldwin

Robert

. 1987. “The Political Market for Protection in Industrial Countries.” In Protection, Cooperation, Integration and Development, edited by El-Agraan

Ali

(pp. 20-37). New York: Macmillan Press.

Baldwin

Robert

Magee

Christopher

. 2000. “Is Trade Policy for Sale? Congressional Voting on Recent Trade Bills.” Public Choice 105:79–101.

Beaulieu

Eugene

Magee

Christopher

. 2004. “Four Simple Tests of Campaign Contributions and Trade Policy Preferences.” Economics & Politics 16:163–87.

Bergstrand

Jeffrey

. 1983. “Measurement and Determinants of Intra-Industry International Trade.” In Imperfect Competition and International Trade, edited by Greenaway

David

Tharakan

P. K. Mathew

, 201–62. Brighton: Wheatsheaf.

Busch

Marc

Reinhardt

Eric

. 2003. “Developing Countries and General Agreement on Tariffs and Trade/World Trade Organization Dispute Settlement.” Journal of World Trade 37:719–36.

Chappell

Henry

. 1982. “Campaign Contributions and Congressional Voting: A Simultaneous Probit-Tobit Model.” The Review of Economics and Statistics 64:77–83.

Cox

Gary

McCubbins

Mathew

. 2005. Setting the Agenda: Responsible Party Government in the U.S. House of Representatives. New York: Cambridge University Press.

Fordham

Benjamin

. 1998. “Economic Interests, Party, and Ideology in Early Cold War Era US Foreign Policy.” International Organization 52:359–96.

10.

Fordham

Benjamin

McKeown

Timothy

. 2003. “Selection and Influence: Interest Groups and Congressional Voting on Trade Policy.” International Organization 57:519–49.

11.

Gartzke

Erik

Wrighton

Mark

. 1998. “Thinking Globally or Acting Locally? Determinants of the GATT Vote in Congress.” Legislative Studies Quarterly 23:33–55.

12.

Gourevitch

Peter

. 1986. Politics in Hard Times: Comparative Responses to International Economic Crises. New York: Cornell University Press.

13.

Grimwade

Migel

. 2000. International Trade: New Patterns of Trade, Production and Investment. Oxon: Routledge.

14.

Grossman

Gene

Helpman

Elhanan

. 1995. “The Politics of Free Trade Agreements.” The American Economic Review 85:667–90.

15.

Henisz

Witold

Mansfield

Edward

. 2006. “Votes and Vetoes: The Political Determinants of Commercial Openness.” International Studies Quarterly 50:189–212.

16.

Hiscox

Michael

. 2002. International Trade and Political Conflict: Commerce, Coalitions, and Mobility. Princeton: Princeton University Press.

17.

Hyejoon

Sung

Hankyoung

. 2011. “Empirical Analyses of US Congressional Voting on Recent FTA.” The B.E. Journal of Economic Analysis & Policy 11:1–35.

18.

Jackson

John

Kingdon

John

. 1992. “Ideology, Interest Group Scores, and Legislative Votes.” American Journal of Political Science 36:805–23.

19.

Jacobson

Gary

. 1990. The Electoral Origins of Divided Government: Competition in US House Election, 1946-1988. Boston: Little, Brown.

20.

Kang

In-Bong

Greene

Kenneth

. 1999. “A Political Economic Analysis of Congressional Voting Patterns on NAFTA.” Public Choice 98:385–97.

21.

Kingdon

John

. 1989. Congressmen’s Voting Decisions. Ann Arbor: University of Michigan Press.

22.

Krehbiel

Keith

. 1996. “Committee Power, Leadership, and the Median Voter: Evidence from the Smoking Ban.” Journal of Law, Economics, & Organization 12:234–56.

23.

Ladewig

Jeffrey

. 2006. “Domestic Influences on International Trade Policy: Factor Mobility in the United States, 1963 to 1992.” International Organization 60:69–103.

24.

Lindsay

James

. 1990. “Parochialism, Policy, and Constituency Constraints: Congressional Voting on Strategic Weapons Systems.” American Journal of Political Science 34:936–60.

25.

Lohmann

Susanne

O’Halloran

Sharyn

. 1994. “Divided Government and US Trade Policy: Theory and Evidence.” International Organization 48:595–632.

26.

Mansfield

Edward

Milner

Helen

Pevehouse

John

. 2008. “Democracy, Veto Players and the Depth of Regional Integration.” The World Economy 31:67–96.

27.

Manyin

Mark

Nikitin

Mary Beth

Chanlett-Avery

Emma

Cooper

William

Rinehart

Ian

. 2014. “US-South Korea Relations.” Congressional Research Service R41481. http://fas.org/sgp/crs/row/R41481.pdf.

28.

Mayda

Maria

Rodrik

Dani

. 2005. “Why Are Some People (and Countries) More Protectionist than Others?” European Economic Review 49:1393–430.

29.

McArthur

John

Marks

Stephen

. 1988. “Constituent Interest vs. Legislator Ideology: The Role of Political Opportunity Cost.” Economic Inquiry 26:461–70.

30.

Midford

Paul

. 1993. “International Trade and Domestic Politics: Improving on Rogowski’s Model of Political Alignments.” International Organization 47:535–64.

31.

Nelson

Douglas

Silberberg

Eugene

. 1987. “Ideology and Legislator Shirking.” Economic Inquiry 25:15–25.

32.

Nollen

Stanley

Quinn

Dennis

. 1994. “Free Trade, Fair Trade, Strategic Trade, and Protectionism in the US Congress, 1987–88.” International Organization 48:491–525.

33.

Oleszek

Walter

. 1984. Congressional Procedures and the Policy process. 2nd ed. Washington, DC: CQ Press.

34.

Poole

Keith

Rosenthal

Howard

. 2001. “D-Nominate after 10 Years: A Comparative Update to Congress: A Political-Economic History of Roll-Call Voting.” Legislative Studies Quarterly 26:5–29.

35.

Putnam

Robert

. 1988. “Diplomacy and Domestic Politics: The Logic of Two-Level Games.” International Organization 42:427–60.

36.

Rogowski

Ronald

. 1989. Commerce and Coalitions: How Trade Affects Domestic Political Alignments. Princeton: Princeton University Press.

37.

Rohde

David

. 1994. “Parties and Committees in the House: Member Motivations, Issues, and Institutional Arrangements.” Legislative Studies Quarterly 19:341–59.

38.

Romer

Thomas

Snyder

James

. 1994. “An Empirical Investigation of the Dynamics of PAC Contributions.” American Journal of Political Science 38:745–69.

39.

Schneider

Howard

. 2010. “Obama, Lee Outlined U.S.-Korea Trade Deal in Seoul, Official Says.” The Washington Post, December 16. http://voices.washingtonpost.com/political-economy/2010/12/post_1.html.

40.

Shepsle

Kenneth

Weingast

Barry

. 1987. “The Institutional Foundations of Committee Power.” The American Political Science Review 81:85–104.

41.

Tosini

Suzanne

Tower

Edward

. 1987. “The Textile Bill of 1985: The Determinants of Congressional Voting Patterns.” Public Choice 54:19–26.

42.

Uslaner

Eric

. 1998. “Let the Chits Fall Where They May? Executive and Constituency Influences on Congressional Voting on NAFTA.” Legislative Studies Quarterly 23:347–71.

43.

Weller

Nicholas

. 2009. “Trading Policy: Constituents and Party in U.S. Trade Policy.” Public Choice 141:87–101.

44.

Williams

Brock

Manyin

Mark

Jurenas

Remy

Platzer

Michaela

. 2014. “The U.S.-South Korea Free Trade Agreement (KORUS FTA): Provisions and Implications.” Congressional Research Service RL34330. http://www.fas.org/sgp/crs/row/RL34330.pdf.

45.

Wink

Kenneth

Livingston

Don

Garand

James

. 1996. “Dispositions, Constituencies, and Cross-Pressures: Modeling Roll-Call Voting on the North American Free Trade Agreement in the US House.” Political Research Quarterly 49:749–70.

46.

Xie

Tao

. 2006. “Congressional Roll Call Voting on China Trade Policy.” American Politics Research 34:732–58.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.08 MB

0.00 MB