Debunking myths about oil: A case study of oil subsidies

Abstract

Despite increasing demands for the reform of oil subsidies, the United States government fails to enact substantial reform policies on the issue. The paper visits the biggest unresolved cleavage in the environmental policy literature where there have been no attempts to quantitatively assess the influence of lobbying and mass participation on the policy-making process. It thus attempts to quantify and examine various factors behind legislators’ votes, and the results are hard to square with a pure lobbying model. While the role of lobbying is certainly not ruled out of the explanatory model per se, this paper observed that congressional preferences may instead also be driven by the voter perception towards environmental regulation in each state. The thrust of the argument is that lobbying, while being a decisive factor, may not be the only one influencing legislators’ decisions for the oil subsidy reform bills. This study hypothesizes that the exchange model theory might not fully provide an explanation of why oil subsidies continuously fall through. It suggests that oil politics may instead follow the neo-pluralist model: While lobbying is an important factor in voting results, legislators are mindful of voters’ perspectives in spite of the fact that they are unorganized—and that they might in fact be even more powerful determinants than the lobby variable.

Keywords

Environmental policy lobbying exchange theory neo-pluralist theory legislative behaviors

In any US presidential election year, it is inevitable that both Republican and Democrat candidates will begin their heated discussion on tax breaks and subsidies. Some vigorously bash the colossal amount of money that corporations walk away with via the so-called “tax breaks,” while others emphasize the indispensability of such subsidies to promote domestic industries.

It is also widely acknowledged that oil industries have much to do with the party disputes; they heavily influence the politics by contributing to the elections in the form of individual, political action committee (PAC)/soft money contributions or by lobbying strategies, with a majority of the recipients being Republicans. This not-so-implicit cozy relationship between the oil industries and the Republicans has continued at least since 20th the century, with many even commenting that the oil industry is one of the few to remain “rock-solid red.”

Notwithstanding the frenzied debate in every election period, oil subsidies never went through any substantial reforms, and big oil companies are continuously reaping the benefits of existing subsidy laws. The media and numerous environmental organizations are thus blaming both the Republicans and the oil industry for pressing on with the policy that the public explicitly disapproves of. They insist that voters are in fact mostly in favor of eliminating tax breaks for the oil industry—and they do have a point. A recent NBC/Wall Street Journal poll found that 74% of US citizens support eliminating tax breaks to oil companies. Then, as everyone seems to implicitly agree, oil policies can only be interpreted as partisan politics. Oil companies have way too much capital in their hands for the environmental organizations or ordinary citizens to exert influence in the process. Oil policies are helplessly in the hands of oil industries and legislators that benefit each other, and the civil society is powerless when it comes to oil issues.

This paper starts with a simple question: Is it true? Are the lobbying and contribution activities of the huge oil industries enough to account for the fact that the reform of oil subsidies in the USA repeatedly fell through? Or were other factors responsible for bringing about such results? More broadly, what is behind this lethargy in US oil policies? This study visits the biggest unresolved cleavage in the literature where there have been no attempts to quantitatively compare the influence of each agent on the domestic oil policy-making process. It attempts to cast light on the motivations guiding state legislators’ oil policy choices by measuring the influence of lobbying and voters’ attitude towards environmental regulation. The study further examines whether the legislators also take other factors into consideration: Legislators’ party, the level of oil industry specialization in each state, environmental expenditures in each state, and gender.

The study thus aims to explain the decisions of legislators via the neo-pluralist model. It pays heed to the current consensus that oil policies follow the exchange model but instead assumes that it might not be the case. According to the exchange model perspective, oil subsidy issues will most likely be heavily swayed by few stakeholders with large interests—in this case the lobbying of big oil companies. By laying out possible alternative factors underlying the failure in oil subsidy reform and conducting quantitative analysis, the study aims to prove the hypothesis that legislators do pay a considerable amount of attention to the constituents. Even with the lobby variable, significance of the voter perception variable indicates the salience of citizens’ interests, regardless of how scattered and unorganized they may be.

Literature review

One of the most important ways in which groups achieve their self-interested goals is through lobbying elected representatives. Numerous papers were invested in identifying this relationship between the lobbying groups and legislative institutions. The literature is largely divided into two branches: One that views lobbying as a form of exchange, and the other that highlights its persuasive role (Hall and Deardorff, 2006). While there are many variants to both approaches, the former assumes that interest group agents and legislators often engage in mutually beneficial exchanges, or “implicit trades,” such as campaign contributions for votes (Morton and Cameron, 1992). This will be explained in depth in the following section.

The latter group argues that legislators are reelection-minded officials who are uncertain about positions to take to be reelected (Bowers, 1992). Therefore, interest groups can use their resources to convince legislators that taking group-friendly positions would be beneficial for their electoral interests. Hansen adopts the notion of “asymmetry of information” (Bowers, 1992). He argues that lobbies are influential because they determine the kinds of information that are available or unavailable to the legislators.

More specifically, the effects of lobbying on environmental policies have also been repeatedly dealt with. Conventionally considered “secondary policies,” extensive studies were conducted on whether both business and environmental interest groups would have a say in legislators’ decisions. Damania and Fredriksson (2003) analyzed the effect of trade liberalization on environmental policy, taking into consideration the collective action problems associated with lobbying. They have also identified that the effect of trade liberalization on the stringency of environmental policy depends on the level of corruption in a country (Damania and Fredriksson, 2003). There have also been attempts to clarify domestic influences on global environmental policies. Sutherland (2001) has identified that the core of lobbying activities on US foreign oil policies consists of American Jews who strive to alter their policies in favor of Israel.

Surprisingly, little systematic research has been conducted solely on the lack of reform. Recently, Pani and Perroni (2018) have identified the policy commitment problem to be a general cause of the lack of energy subsidy reforms, focusing on the incentive of rational policy-makers. Even when reforming energy subsidy is feasible, policy-makers may deliberately refrain from doing so due to incentives for reelection. Yet, it does not fully explain lobby influence on the domestic oil industry specifically. This research attempts to narrow its scope down to oil subsidy bills, while at the same time testing the strength of its influence on legislators.

Subsidy

According to the World Trade Organization (WTO) definition, subsidies include: (a) all government financial contributions or direct support; (b) transfer of risk through provision of debt, equity, and guarantees; (c) forgone revenue through tax breaks; (d) provision of infrastructure, goods, and services below market value; and (e) royalty breaks and investment in infrastructure (WTO Agreement, 1867). As for the fossil fuel subsidy, the International Energy Agency defined it in 2006 as “any government action that lowers the cost of fossil fuel energy production, raises the price received by energy producers or lowers the price paid by energy consumers (Sutherland, 2001).” Current US subsidies to oil industries are mostly focused on production subsidies, thereby benefiting the oil producers to foster domestic industry.

Recently, there has been a surge in research that pinpoints the inefficiency associated with oil subsidies, but most of the literature centers on developing countries and their governments’ incompetence in reforming the long-lasting practice. Benefits of removing oil subsidies are also regularly dealt with. China’s OECD report on the USA confirms that energy security no longer justifies the notion of supporting the producers, since the USA has already considerably increased its domestic production capacity (OECD, 2016). Moreover, according to the report, the current oil subsidies are giving oil and gas companies a competitive edge over any other types of energy companies, thereby hindering their development.

Then why do subsidies persist? Most of the previous literature on subsidy reforms focuses on the study of energy subsidies in general. Such research acknowledges that energy subsidies are in need of reform, and that governments fail to implement such changes despite the obvious benefits to their economy. Whitley and Van der Burg (2015) have identified several reasons behind this persistence of subsidies in many countries around the world:

Historically, national governments have controlled the production, price, and value of energy industries for their benefit.

There is limited transparent information about provided subsidies in general.

There are many misperceptions about the effectiveness of fossil fuel subsidies in economic development.

Subsidies are often vigorously defended by special interests because of the benefits they bring to specific sectors and groups.

Especially for the USA, the primary reason for the continuation of oil subsidies rests upon the interests of big oil companies, and the fact that the fossil fuel industries often have access to many levels and branches of the government. Many works are dedicated to uncovering the business strategies in financing PAC contributions. However, it must be emphasized that policy is not a straightforward reflection of private interests into government policies. Legislators are the ultimate decision-makers, not the oil industries; therefore, in-depth analyses of politicians’ selection of choices and what influences their decisions are obligatory.

Legislators as decision-makers

Legislators, as “office seekers,” aim to maximize the votes they receive in the following election. Legislators can do so by providing services to interest groups, or take into account the views of the public when enacting a bill. Previous literature on the policy-making process assumes three sets of agents: Voters, interest groups, and legislators. Policy outcomes depend on the comparative advantages of these groups; in other words what each agent has to offer the others (Denzau and Munger, 1986). Bawn and Thies (2003) have focused on their behavior under a certain legislative system, for instance that they tend to be more responsive to interest groups under the proportional representation (PR) system. Others attempted to come up with theories that could systematically analyze the reasons behind legislators’ decisions.

The “elite theory” demonstrates that a small number of individuals in powerful positions dominate the legislative process. Though partially convincing, the theory is inconsistent with various policies that benefit the minorities in society. In need of a better explanation, another type of literature puts forward a “pluralist theory.” A set of scholars including Robert Dahl argues that election coerces the legislators to be accountable not only to the “elites” but also to the preferences of voters. Since the voters have the rights to select their representatives via election, it forces policy-makers to take into consideration the policies that voters desire. Dahl (1961) insists that “few individuals are without access” to any resources useful in pursuing influence; therefore, they can easily join interest organizations that they support.

However, over the past several decades, many have questioned the validity of the theory. Even inside the pluralist school, doubts have risen about its ability to explain the US system, or about its tendency to downplay the class system (Manley, 1983). Some identified possible limitations of the theory, pointing out that a rational person would attempt to free-ride on the efforts of others as much as possible. In other words, individuals with low stake in a certain policy outcome are unlikely to organize or take part in organizations and spend their resources. An alternative “exchange model” was developed by Olson (2009). The theory holds that a large number of people with small stakes would fail to have collective mind to affect the legislative process. Consequently, a few individuals with a large stake will defeat “the many.” Olson’s logic of collective action therefore predicted that lobbies for mass group of individuals would not exist unless they provided members with benefits worth more than the cost of joining.

Along a similar line, Lowi (1964) argues that the design of America’s political institutions hurts “the many.” Only the small, wealthy producer groups will have substantive advantage in politics as they can make better “exchanges” with the legislators, leaving many other interests unrepresented. Some policy areas might be controlled by a narrow coalition seeking to direct the distribution of discrete benefits by government, such as the purchase of expensive military aircraft (McFarland, 2004). US public policy-making is thus infused with particularistic policy coalitions, forming multiple elites within the system. Lowi (1964) further develops the argument into discussing subsidy policies in general. According to him, most subsidy policies are enacted in political sub-systems named the “iron triangles.” They consist of self-interested producer groups, bureaucratic officials, and legislators of congress who form a certain coalition to dominate specific policy areas. In this process, legislators are granted their freedom of choice entirely in the process, yet decisions are made largely in favor of a small number of individuals.

Yet another group of scholars strived to mediate the two opposing views. Stemming from the pluralist theory is the “neo-pluralist theory.” As could be assumed by its name, neo-pluralist theory evolved through discarding while modifying the pluralist theory. Its proponents understood that proliferation of groups does not necessarily imply a fair and representative policy-making process in some issue domain. Thus, while pluralist in its acknowledgment of diverse interest representation in US politics, it is “neo” in that the theory understands that the diversity could be constrained by a number of factors. Still, it claims that although interest groups all have different levels of available resources, they mostly receive a fair amount of representation and can compete under a democratic environment. For neo-pluralists, if mass opinion effectively constrains policy at the broadest levels, then organized interests can change policy only so far and so fast (Lowery and Gray, 2004). Smith (2000) also notes that a powerful business interest is readily checked when the policy issue at stake is salient and public opinion runs counter to business interests.

Main argument

The thrust of the argument is that while both narrow coalitions of small stakeholders with large interests and mass groups of individuals with scattered interests affect legislators’ voting results towards environmental bills, legislators and their decisions are also likely to be swayed by citizens’ attitudes towards environmental regulation. In other words, environmental policy-making can be understood with the neo-pluralist model. It is pluralistic in the sense that individuals have a definite say in the environmental policy-making process, but it also distinctively follows the “neo” pluralistic theory in that lobbying is also a decisive factor in influencing legislators’ decisions on the oil subsidy reform bills.

Previous research mostly focuses on the respective influences of both lobbying and citizens on the enactment of economic policies. In other words, while they argue how either the power of money or the power of mass perception affects legislative policy-making, there was no research that examined the validity of electorates’ perceptions even after taking into account the influence of lobbying on the legislators. Along the same line, articles that emphasize public participation in the environmental sphere mostly focus on uncovering institutional mechanisms for allowing the lay public to influence environmental risk decisions (Fiorino, 1990), or designing participatory processes that could effectively integrate public input into the creation of environmental policies (Renn, 2006). This research, motivated by a lack of research that compares the degree of pressures that both large-scale firms and mass constituents levy on legislators and their decisions, seeks to quantitatively compare the two and, in a broader sense, uncover the process behind legislative policy-making on environmental issues.

This study thus hypothesizes that the exchange model theory might not fully provide an explanation of oil subsidy reform bills. It suggests that oil politics follow the neo-pluralist model: Legislators are mindful of voters’ perspectives in spite of the fact that they are unorganized. We therefore propose a neo-pluralist perspective as a more suitable framework for examining and explaining oil subsidy issues in the USA. In other words, we argue that both the lobbying of big oil companies and the mass opinion towards environmental regulations are effective indicators of legislative decisions. The study’s hypotheses are thus as follows:

H1: If a legislator receives more lobbying related to oil and gas issues, the legislator is more likely to vote against oil subsidy reform.

H2: If larger number of voters prefer strict environmental laws and regulations, a legislator is more likely to vote for oil subsidy reform.

H3: Voter perception towards environmental regulations has better explanatory power than the amount of lobbying.

Data analysis

This study conducts a quantitative analysis, more specifically regression analysis, to examine the significance and relative influence of the independent variables. Opposition rate of the oil subsidy reform bills for each state is calculated by analyzing a total of four oil subsidy reform bills, three from the House of Representatives (year 2011, 2014, and 2015) and one from Senate (2011). Most of the existing literature measures the effect of private money on public policy via roll-call votes of the legislators. The paper adopts the methodology and looks into each voting decision of respective legislators. Within the four bills, there are in total 1322 observations of legislator-bill case. This study will first describe each variable used in the analysis and demonstrate the result of the analysis. Summary statistics are attached in the Appendix, Table A1.

Dependent variable

The dependent variable of this paper is the binary outcome of such decision-making: The YES votes indicate a pro-environmental decision—arguing for the elimination of oil subsidies, while the NO votes point otherwise. In the dataset, the former case is coded as 1, and the latter as 0. Brief descriptions of each bill are given:

a. House 2011: Offshore drilling subsidies

Ed Markey (D-MA) offered the amendment to H.R. 1, the Full-Year Continuing Appropriations Act of 2011, to eliminate up to US$53 billion in taxpayer subsidies by closing a royalty payment loophole for oil companies operating offshore. On February 18, the House rejected the amendment by a vote of 174 to 251.

b. House 2014: Offshore drilling subsidies

Representative Earl Blumenauer (D-OR) offered an amendment to H.R. 4899, the Lowering Gasoline Prices to Fuel an America That Works Act of 2014, which would require companies to renegotiate leases that require no payment of royalties to the USA before bidding on new leases authorized by the underlying bill. On June 26, the House rejected the Blumenauer amendment by a vote of 179 to 229.

c. House 2015: Big oil subsidies on public lands

Representative Steve Pearce (R-NM) offered an amendment to H.R. 2822, the Department of the Interior, Environment, and Related Agencies Appropriations Act, 2016, which would block any efforts to raise the royalty rates on oil and gas produced on onshore federal public lands. On July 8, the Pearce amendment was approved by a vote of 231 to 198.

d. Senate 2011: Oil subsidies

Senator Robert Menendez (D-NJ) sponsored S. 940, the Close Big Oil Tax Loopholes Act. It would close tax loopholes for oil companies, under the argument that their outsized profits call for reduced subsidies. On May 17, the Senate failed to achieve the 60-vote threshold needed to pass a motion to proceed to the consideration of S. 940. The motion failed on a 52 to 48 vote.

Independent variable

One of the key pillars of lobbying’s effectiveness is its influence in the US Congress. An extensive literature attempts to measure the political efficacy of interest group donations. Most of them examine the effects of contributions on roll call votes by the members of Congress. As mentioned above in depth, interest groups are intense demanders in the policy-making process who utilize their resources to persuade. This study acknowledges that oil companies, with their unmatched supply of finance, bring about various methods to affect the political sphere and secure their interests. Thus, it hypothesizes in line with the exchange theory that lobbying activities will indeed sway legislators’ decisions. The LOBBY variable is extracted from opensecrets.com, and the actual amount of money to each legislator for the respective election cycle is reported in US$1000. It is expected that there would be a negative relationship between lobbying and legislators’ votes for oil subsidy reform.

On the other hand, the main independent variable in the second hypothesis is voter perception towards environmental regulation. Voters are a heterogeneous, unorganized group of people with diverse interests, thus making it hard for the legislators to accommodate all perspectives. Yet, it is also a long-held belief that public officials should, and do, represent the voters in their respective constituencies. Taking note of the neo-pluralist theory of interest realization, this study postulates that voters’ opinion on environmental regulation would significantly matter to the legislators when making judgment calls for the oil subsidy reform bills.

The opinions of voters towards stricter environmental regulations come from the Pew Research Survey conducted in each year of the bill, asking the respondents whether:

stricter environmental laws and regulations cost too many jobs and hurt the economy;

stricter environmental laws and regulations are worth the cost.

The strong tone in the question indicates that those who answered positively have a certain degree of commitment towards strict environmental regulations, thereby allowing us to assume that environmental issues would be one of the factors they include in their voting behaviors. The variable PERCEPTION is coded as the percentage of respondents who chose the second option as their answer; in other words, the ratio of voters, who are from the same state of a legislator, in favor of pro-environmental legislation. It should be noted, however, that the data represents voters’ overall support for strict environmental regulations rather than their perception towards the oil industry. Nevertheless, the study decided that it is possible to draw the inference between tendencies to argue for harsh regulations even with economic costs and their support for oil subsidy reform. This paper expects to observe a positive relationship between voter perception and voting decision.

Other than the main explanatory variables, the study also attends to the independent variables expected to have a certain effect of each on the study’s dependent variables: Factors belonging to an individual legislator and factors belonging to a state of a legislator. For the former, this study includes a legislator’s party affiliation, term, and gender. Then for the latter, a state’s basic economic circumstances (GDP growth rate, unemployment rate, and inflation), oil industry specialization rate, and government expenditure on environmental issues are taken into consideration.

To begin with, numerous papers have argued the strong effect of legislators’ party affiliations on environmental policy decisions. While some scholars limit the scope to local congressional representatives, others have expanded their range to federal levels as well. Some attempted to uncover the relationship between legislators’ party affiliation and their decisions regarding environmental policies (Innes and Mitra, 2015). Taking such a vast amount of literature into account, this paper assumes that the party affiliation of legislators would serve as an indicator for a Congressional representative’s oil subsidy preferences. In other words, Democratic partisanship might influence the legislators’ inclination to vote for the oil subsidy reform bills. The PARTY variable of the dataset is coded 1 if the legislator belongs to the Democratic party and coded 0 if the legislator belongs to the Republican party. According to conventional prediction, there would be a positive relationship between party identification and the dependent variable.

Furthermore, the term of a legislator can indirectly indicate a degree of electoral competition. It is intuitive to assume that the longer a term of a legislator is, the more stable position he or she is in regarding the electoral competition. On the other hand, a shorter term demonstrates that a legislator has not yet taken full advantage at the competition. It would be easier for voters to sway to another candidate from a Congressional representative with a shorter history. The TERM variable of the dataset is the count of the number of terms of the legislator, including the current cycle.

Finally, female House members are more likely to favor stricter environmental policies than male members (Fredriksson et al., 2011). Mandel and Dodson (1993), Mezey (1994), Sapiro (1981), and Thomas (1991) also found similar results, arguing that female elected officials will produce more liberal public policy. Thus, drawing from past researches, this paper postulates a significant relationship between gender and voting records of oil subsidy bills: Women legislators will be more likely to vote in favor of the oil subsidy reform bills. The GENDER variable was added into the regression model as binary variables. Value 1 is given if a legislator is male, and 0 if female, indicating that there would be negative relationship between gender variable and dependent variable.

The most fundamental indexes of an economic situation are GDP growth rate, unemployment rate, and inflation. All three are included, yet this study is specifically related to the oil subsidy issue. Dell (2009) and Portney (2002) have asserted that legislators take economic factors into account when voting for green energy laws. Dell finds that extractive industry strength has a negative effect on the adoption of pro-environmental laws. Along the same line of reasoning, Portney examined the sustainability initiatives of 24 cities. He argues that cities that rely less on manufacturing industries tend to be more supportive of sustainability initiatives. Thus, this study contends that the oil industry would also be intricately linked with economic interests. Since state legislators would tend to pursue policies that would contribute to the state economy, the paper infers a positive relationship between a state’s gains from the oil industry and the opposition rate for oil subsidy reform bills.

The Bureau of Economic Analysis’ Industry Specialization Index (ISI) is a measure of the degree to which states are specialized in an industry. The more specialized a state is in an industry, the higher the number will be. Looking into the ISI of the oil industry in each state might reveal another alternative factor for the legislators’ behaviors. The oil industry specialization rate of each state was gathered from the Bureau of Economic Analysis. Consequently, this control variable is likely to have a negative correlation with the dependent variable.

Each state’s degree of environmental regulation is included in the model as a control variable. The underlying reasoning is that the policy environment indicating the overall environmental stringency of states may have a sizable influence on state legislators’ decisions on environment-related policies. Level of regulation is measured by dividing the state environmental budget by its GDP. Of course, there are certain limitations to this measurement, which have been mentioned enough in previous literature—mostly that the data fails to incorporate any local efforts that could not pass muster in the state budgetary system. Regulatory economies of scale could also be ignored. Therefore, states with large budgets could be misclassified as less enthusiastic towards environmental regulations (List and Co, 2000).

Kahn (2002) nevertheless adopts this method to measure how local governments’ allocation of budget expenditure on environmental matters binds the state’s overall attitude towards environmentalism. Therefore, this study also maintains that assembling data on state environmental expenditure will be an effective quantification of states and how much they take initiative in protecting the environment. In terms of collecting the data, state environmental expenditures are gathered from ballotpedia.com. The data includes the funds in each state legislature that go toward local agencies related to environmental protection between 2011 and 2015. The figures are divided by the GDP rate of each state for the respective year. Then, the expectation is that higher environmental expenditures in a legislator’s state would lead to his or her more favorable voting for oil subsidy reform.

Results

The study produced four logistic regression models. Model 1 only includes LOBBY as the independent variable, testing the first hypothesis. Model 2 replaces the independent variable with PERCEPTION in order to test the second hypothesis. Then, Model 3 includes both independent variables, making the comparison between two hypotheses more feasible. Lastly, Model 4 take into account the interactive effect between either independent variable and TERM in order to introduce more elaborate control to the analysis. Table 1 shows the results of all four models and Figure 1 presents line plots representing each effect of lobbying and voter perception grounded on the results from Model 3. Figure 2 presents line plots representing each marginal effect conditioned by TERM.

Table 1.

Effect of lobbying and voter perception on voting for oil subsidy reform.

	Dependent variable
	VOTE
	(1)	(2)	(3)	(4)
LOBBY	−0.019***		−0.020***	−0.010
LOBBY	(0.005)		(0.005)	(0.007)
LOBBYTERM*				−0.001*
LOBBYTERM*				(0.001)
PERCEPTION		6.076**	5.063**	4.688*
PERCEPTION		(2.490)	(2.498)	(2.522)
PERCEPTIONTERM*				0.148***
PERCEPTIONTERM*				(0.053)
PARTY	5.643***	5.887***	5.666***	5.676***
PARTY	(0.290)	(0.297)	(0.299)	(0.300)
GENDER	−0.787**	−0.964***	−0.794**	−0.805**
GENDER	(0.376)	(0.364)	(0.380)	(0.381)
GDP	−0.049	−0.040	−0.041	−0.038
GDP	(0.034)	(0.035)	(0.035)	(0.035)
UNEMPLOYMENT	−0.083*	−0.043	−0.066	−0.062
UNEMPLOYMENT	(0.046)	(0.047)	(0.048)	(0.048)
INFLATION	0.097***	0.093***	0.087***	0.085***
INFLATION	(0.028)	(0.029)	(0.029)	(0.029)
INDUSTRY	−0.250***	−0.362***	−0.254***	−0.239***
INDUSTRY	(0.075)	(0.070)	(0.077)	(0.079)
EXPENDITURE	−0.111	−0.124	−0.080	−0.075
EXPENDITURE	(0.150)	(0.149)	(0.151)	(0.150)
CONSTANT	−5.415***	−11.066***	−8.732***	−8.833***
CONSTANT	(2.447)	(3.047)	(3.111)	(3.138)
N	1,322	1,316	1,316	1,316
Log likelihood	−232.185	−238.443	−229.591	−225.618

Note: *p<0.1; **p<0.05; ***p<0.01.

Figure 1.

Visualized effects of lobbying and perception on probability.

Figure 2.

Marginal effects of lobbying and perception conditioned by term.

According to the results presented in Table 1, independent variable of LOBBY does demonstrate negative and significant coefficients on the dependent variable of VOTE in Model 1, which supports the first hypothesis. Lobbying clearly has an effect on the behavior of legislators: As a legislator receives more lobbying from oil and gas-related interest groups, the legislator is more likely to vote against oil subsidy reform, favoring those interest groups in return.

Thus, the idea of exchange theorists who believe that interest groups wield power over the legislators’ decisions is in fact plausible. However, their depiction of the pluralist view as inaccurate and naïve is wrong at the same time. From the results of Model 2, voter perception was verified to be another plausible explanation for the continuation of oil subsidies; legislators are mindful of the voters’ opinion on environmental regulations in their constituencies. Individual members of the community, despite being unorganized, did notably affect the Congressional representatives and their voting decisions, supporting the second hypothesis in the end.

The findings regarding the two main independent variables are consistent in Model 3 as well. Including both independent variables in the multivariate logistic regression model does not result in a serious change in any relationship. LOBBY still has a negative and significant relationship with VOTE and, conversely, PERCEPTION still has a positive and significant relationship with VOTE. Neither had any serious impact against each other, meaning that each hypothesis is well supported discretely.

Interestingly, other control variables also presented consistent result in all three models, mostly following the expectation above. The PARTY variable was undoubtedly one of the strongest explanatory factors of voting decisions, adding validity to the conventional wisdom that oil politics is indeed partisan. Republican legislators were likely to vote against the oil subsidy reform bills. The political deadlock of oil subsidy bills is proven to be largely due to rigid political institutions. INDUSTRY was also negative and significant. On the other hand, basic economic variables were mostly insignificant except for INFLATION which consistently demonstrated a positive and significant effect. Legislators are fact dependent on economic conditions, but more specifically whether the state economy depends on the oil industry. Last, GENDER was also pertinent in the voting decisions. As the previous literature indicated, female legislators displayed higher tendencies to vote for the reform bills. EXPENDITURE did not bring the expected result, showing statistical insignificance in all three models.

Comparing the coefficients in a logistic regression model is not necessarily an accurate method to determine the substantial superiority of effects between two different independent variables. However, this paper does have an interest in clarifying which of the two theories (exchange theory and neo-pluralist theory) hold more explanatory power in the real world. Therefore, the paper provides the change of probability to vote in favor of pro-environment legislation affected by each independent variable, partially allowing rather fair comparison between two hypotheses. In the two graphs shown in Figure 1, the absolute term of slope of LOBBY changes more radically than that of PERCEPTION. In other words, it can be concluded that the explanation by neo-pluralist theory demonstrates its substantial effect more strongly than, or at least as strongly as, exchange theory. Table 2 demonstrates the probability and its change. Although the maximum value of LOBBY is a lot bigger than US$400,000, the probability reaches 0 at that point, so the report regarding LOBBY ends at the value of 400.

Table 2.

Predicted effects of lobbying and perception on probability.

LOBBY	Probability	Change	PERCEPTION	Probability	Change
0	0.515		0.000	0.033
100	0.129	−0.386	0.250	0.107	+0.074
200	0.020	−0.109	0.500	0.297	+0.190
300	0.003	−0.071	0.750	0.600	+0.303
400	0.000	−0.003	1.000	0.842	+0.242

Model 4 introduces two interaction terms: Interaction between LOBBY and TERM, and interaction between PERCEPTION and TERM. The main purpose is to control the degree of competition and stability of a legislator’s position. However, TERM itself cannot theoretically affect whether a legislator would be in favor of a pro-environment bill or not. It is reasonable to adopt more deliberate methodology for the interaction term in order to analyze the indirect effect of TERM on a legislator’s sensitivity to more directly influential variables, which are LOBBY and PERCEPTION. According to the results, the interaction term of LOBBY is not significant, indicating that the length of term does not influence a legislator’s sensitivity to interest groups’ lobbying. Conversely, the interaction term of PERCEPTION is significant. In other words, the length of term does influence a legislator’s sensitivity to voters’ perception. Moreover, the direction of the term is identical to that of the simple effect of PERCEPTION, meaning that the effect of voters’ perception intensifies as the term of a legislator increases. To be more accurate, the visualization of the result is necessary when interpreting the interaction term.

Figure 2 presents the plots of marginal effects of each independent variable conditioned by the term of a legislator. The x-axis in both plots indicates the value of TERM in order to demonstrate how the degree of the independent variable’s effect changes along with the change of TERM. Neither plot largely coincides with the dotted line of zero value, so it can be concluded that the effect is robust with the control of TERM variable. However, in the case of LOBBY, no significantly conditioning effect of TERM could be discovered while the increase of TERM led to a stronger effect of PERCEPTION on VOTE. This indicates that term does not matter for a legislator regarding the effectiveness of lobbying. However, a legislator would be even more likely to be sensitive to voters’ perception if his or her term is longer. Such a finding is quite surprising since it is more intuitive to conclude that voters’ perception may not matter to a legislator who has already seized a stable position in the electoral competition.

Conclusion

Political science is largely the study of power, but power itself is difficult to measure. This paper attempted to quantify the influence of power that each actor in a society can exert in the policy-making process. The analysis came up with rather counterintuitive results. The results are hard to square with a pure lobbying model. While the role of lobbying is certainly not ruled out of the explanatory model per se, this paper observed that congressional preferences may instead be more driven by the pursuit of reelection.

More specifically, although the “money buys influence” argument is certainly plausible, voters’ perception towards environmental regulation turned out to be a stronger signifier for explaining the opposing votes of legislators in oil subsidy reform bills. Contrary to the normal belief that the majority of the voters are in favor of the reform while legislators act against them, the research indicates that legislators are mindful of public opinion.

The pluralist model is first rejected in the analysis with the lobby variable turning out to be one of the most significant indicators of legislative voting behaviors towards oil subsidy reform bills. At the same time, the exchange model, strictly claiming that only small interest groups with high stakes and affluent resources would influence the legislators’ votes, was thus partially rejected. The LOBBY variable does play a big role in forming the “No” votes of Congressional representatives. However, regardless of the lobby amount, states with voters who rigorously give weight to environmental protection nevertheless argued for the oil subsidy reform bills. Oil politics thus follow the neo-pluralist idea of interest representation where public opinion can effectively cancel out the rigidity of “iron triangles.”

To sum up, this study contends that oil subsidy issues act in accordance with the neo-pluralist model in a large fraction. The public’s attitude towards stricter environmental regulations indeed constitutes the strongest evidence of legislator behaviors. Furthermore, the paper sheds light on the possibility to overcome the political gridlock of US oil policies. If voters in each constituency perceive the necessity of oil companies “to pay their fair share,” the votes of legislators could change.

Finally, several limitations of the paper present possible ideas for future research. As mentioned above, the study utilized the voters’ overall support for strict environmental regulations as a measure for public perception instead of their support towards the oil industry specifically due to non-availability of the data. Further research could improve the validity of the work by including more specific variables in the study. Moreover, discrepancies in perceptions of voters in respective states do not change the fact that the majority of citizens in the USA oppose the oil subsidy policy. The institutional limitations that obstruct the majority opinion from being brought into effect should be further examined.

Footnotes

Appendix

Table A1.

Summary statistics of variables.

	Min	q.1	Median	q.3	Max	Mean	SD
VOTE	0.000	0.000	0.000	1.000	1.000	0.483	0.500
LOBBY	0.000	0.800	7.100	28.550	955.575	28.285	63.329
PERCEPTION	0.294	0.542	0.571	0.612	0.773	0.572	0.056
PARTY	0.000	0.000	0.000	1.000	1.000	0.472	0.499
TERM	1.000	2.000	5.000	10.000	31.000	6.319	5.113
GENDER	0.000	1.000	1.000	1.000	1.000	0.829	0.376
GDP	33.158	45.440	52.364	57.928	76.225	51.787	8.546
UNEMPLOYMENT	85.000	93.100	99.100	108.600	122.300	100.650	9.522
INFLATION	17.503	40.301	41.940	43.797	51.036	41.648	3.940
INDUSTRY	0.000	0.000	0.020	0.610	18.930	0.982	2.194
EXPENDITURE	0.000	0.480	0.858	1.589	6.761	1.264	1.102

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: funded by Social Science Korea (SSK).

References

Bawn

Thies

(2003) A comparative theory of electoral incentives: Representing the unorganized under PR, plurality and mixed-member electoral systems. Journal of Theoretical Politics 15(1): 5–32.

Bowers

(1992) Gaining access: Congress and the farm lobby, 1919–1981. By Hansen

John Mark

. The Journal of American History 79(3): 1218–1219.

Dahl

(1961) Who Governs? New Haven: Yale University Press.

Damania

Fredriksson

(2003) Trade policy reform, endogenous lobby group formation, and environmental policy. Journal of Economic Behavior & Organization 52(1): 47–69.

Dell

(2009) The Grassroots are greener: Democratic participation and environmental policies in state politics. Review of Policy Research 26(6): 699–727.

Denzau

Munger

(1986) Legislators and interest groups: How unorganized interests get represented. American Political Science Review 80(01): 89–106.

Fiorino

(1990) Citizen participation and environmental risk: A survey of institutional mechanisms. Science, Technology, & Human Values 15(2): 226–243.

Fredriksson

Wang

Mamun

(2011) Are politicians office or policy motivated? The case of U.S. governors’ environmental policies. Journal of Environmental Economics and Management 62(2): 241–253.

Hall

Deardorff

(2006) Lobbying as legislative subsidy. American Political Science Review 100(01): 69–84.

10.

Innes

Mitra

(2014) Parties, politics, and regulation: Evidence from clean air act enforcement. Economic Inquiry 53(1): 522–539.

11.

Kahn

(2002) Demographic change and the demand for environmental regulation. Journal of Policy Analysis and Management: The Journal of the Association for Public Policy Analysis and Management 21(1): 45–62.

12.

List

(2000) The effects of environmental regulations on foreign direct investment. Journal of Environmental Economics and Management 40(1): 1–20.

13.

Lowery

Gray

(2004) A neopluralist perspective on research on organized interests. Political Research Quarterly 57(1): 163.

14.

Lowi

(1964) American business, public policy, case-studies, and political theory. World Politics 16(4): 677–715.

15.

McFarland

(2004) Neopluralism: The Evolution of Political Process Theory. Univ Pr of Kansas.

16.

Manley

(1983) Neo-pluralism: A class analysis of pluralism I and pluralism II. American Political Science Review 77(2): 368–383.

17.

Morton

Cameron

(1992) Elections and the theory of campaign contributions: A survey and critical analysis. Economics & Politics 4(1): 79–108.

18.

OECD (2016) G20 voluntary peer review by China and the United States on inefficient fossil fuel subsidies that encourage wasteful consumption. Report. Available at: http://www.oecd.org/site/tadffss/publication/ (accessed 18 March 2019).

19.

Olson

(2009) The Logic of Collective Action. Cambridge, MA: Harvard University Press.

20.

Portney

(2002) Taking sustainable cities seriously: A comparative analysis of twenty-four US cities. Local Environment 7(4): 363–380.

21.

Renn

(2006) Participatory processes for designing environmental policies. Land Use Policy 23(1): 34–43.

22.

Sapiro

(1981) Research frontier essay: When are interests interesting? The problem of political representation of women. American Political Science Review 75(3): 701–716.

23.

Smith

(2000) American Business and Political Power: Public Opinion, Elections, and Democracy. Chicago: The University of Chicago Press.

24.

Sutherland

(2001) “Big Oil” at the Public Trough? An Examination of Petroleum Subsidies. Policy Analysis 390. Available at: https://www.cato.org/publications/policy-analysis/big-oil-public-trough-examination-petroleum-subsidies (accessed 4 August 2018).

25.

Thomas

(1991) The impact of women on state legislative policies. The Journal of Politics 53(4): 958–976.

26.

Whitley

Van der Burg

(2015) Fossil fuel subsidy reform: From rhetoric to reality. Report, November 2015. Available at: https://www.odi.org/publications/9613-fossil-fuel-subsidy-reform-rhetoric-reality (accessed 5 August 2018).

27.

WTO Agreement (1867) Agreement on Subsidies and Countervailing Measures. Available at: https://treaties.un.org/doc/publication/unts/volume%201867/volume-1867-i-31874-english.pdf (accessed 18 March 2019).