Abstract
Earmarks are an important tool of a legislators’ reelection strategy, as they provide a prime opportunity to benefit their district and increase their chances at reelection. Yet, some legislators secure substantially more earmarks than others. Our study offers a new explanation for why this occurs. With limited legislative resources at their disposal, members of Congress are forced to make trade-offs in how they allocate their resources among legislative activities. In a world with resource constraints, members less engaged in earmarking are likely engaged in policymaking activities. We take advantage of new earmark and policymaking data to explore this hypothesized negative relationship between earmarking and policymaking effort. We find that legislators who exert greater effort on policymaking secure significantly fewer earmark dollars.
Introduction
Appropriation earmarks have long been a common element of many legislators’ Washington strategies. With an eye toward reelection, the logic of pursuing earmarks is straightforward. Earmarks allow publicity seeking legislators to claim credit for the funding of popular local projects, which in turn boosts their recognition and favorability (Grimmer, 2013; Mayhew, 1974). But securing earmarks is not costless. Members of Congress (MCs) must employ nonnegligible amounts of resource and effort to pursue earmarks or other legislative activities. Time and congressional staff resources are finite and impose practical restrictions on the activities of MCs. The constraints imposed by limited legislative resources force MCs to make trade-offs in the degree of effort dedicated to other types of legislative activities, such as oversight, policymaking, or fundraising. The primary consequence of legislators’ finite resources is that effort in one arena of legislating may come at the expense of realizing other opportunities.
Traditional explanations of earmark procurement center around the institutional position of MCs (Engstrom & Vanberg, 2010; Lee, 2005). Although theories of distributive politics centering on institutional rank perform well in explaining the sizable differences in earmarking success between majority and minority status MCs, as well as chairmen and backbenchers, they offer little theoretical insight into the significant variation in earmarking within the party caucus. Despite the power of seniority and of majority party status (Balla, Lawrence, Maltzman, & Sigelman, 2002; Levitt & Snyder, 1995), some MCs well positioned to secure earmarks receive far fewer than their rank would predict. Credit claiming under constrained resources helps explain the discrepancy. The pursuit of earmarks and other legislative opportunities, such as policymaking, is not costless.
To illustrate this theory of legislative trade-offs, we track MCs’ success in the earmarking process against an alternative legislative activity: policymaking. We enlist a dataset of earmarks from the Office of Management and Budget (OMB) that directly identifies sponsors of an earmarked project as well as a novel measure of policymaking effort developed by J. H. Fowler (2006a). Rather than rely on members’ legislative output as a measure of effort, a measure highly endogenous to seniority and majority party status, our policymaking effort statistic measures MCs level of interconnectedness in bill cosponsorship developed from cultivating legislative coalitions in support of their preferred policies. This measure of policymaking effort does not directly suffer from the endogenous advantages associated with majority party status. Our analysis shows a statistically and substantively significant inverse relationship between earmarking success and proactive engagement in the policymaking sphere. Put simply, some MCs work harder to secure earmarks whereas others are more dedicated to the policymaking process. The result is robust to a variety of alternative policymaking actions that cannot be readily restricted by the majority’s agenda setting power, including an MC’s number of bills introduced, as well as the number of floor speeches. Although a change to majority party status still produces the largest marginal increase in earmark dollars, the magnitude of change for 25th percentile to 75th percentile shift in policymaking effort is greater than the change associated with a shift in any other predictor.
Past Literature
Past research explaining variation in earmark procurement largely focuses on how institutional biases and district preferences position certain MCs to secure a greater number of earmarks (Engstrom & Vanberg, 2010; Lee, 2005; Shepsle & Weingast, 1981; Weingast, 1979). Members of the Appropriations Committee (Ferejohn, 1974), as well as more senior members, tend to receive more earmarks than their colleagues (Engstrom & Vanberg, 2010; Ferejohn, 1974; Lee, 2005). However, despite the distributive benefits of seniority, longer tenured members do not secure a greater allocation of discretionary federal spending (A. Fowler & Hall, 2015). 1 Members representing districts ideologically predisposed to disagree with earmarked spending projects seek and receive fewer earmarks (Alvarez & Saving, 1997).
Theoretical and empirical work also find that political parties strategically dole out pork to maximize electoral opportunities (Ferejohn, 1974). Majority party status both predicts greater earmark allotments (Engstrom & Vanberg, 2010), as well as a corresponding bump in electoral performance upon their receipt (Balla et al., 2002; Crespin & Finocchiaro, 2013). More broadly, the total amount of discretionary federal spending is higher in a legislator’s district if they are a member of the majority party (Lee, 2005; Levitt & Snyder, 1995; Rundquist & Carsey, 2002). Earmarks may not be the exclusive mechanism for MCs to steer money toward their districts, but it is the most visible.
Partisan distribution of earmarks is another way in which the majority party takes care of its marginal members (Cox & McCubbins, 1993; Rohde, 1991; Sinclair, 2007). Jenkins and Monroe (2012) show that donations by party leaders’ support party members most ideologically predisposed to coordinate with the minority. It is unsurprising to then find that moderate members are more likely to highlight particularized spending projects secured for their district, perhaps allocated for their loyalty on party votes (Grimmer, 2013). Although widely disparaged by both voters and politicians alike, earmarks remain popular when in voters’ own districts, which explains why MCs want them as a means to increase their reelection chances (Frisch, 1998; Hibbing & Theiss-Morse, 1995). Party leaders take advantage of this relationship and often use earmarks to gain legislators’ support for the passage of legislation (Evans, 2004).
The extant literature paints a picture where most legislators seek earmarks and receive them. Although parties strategically allocate earmarks to their coalitions’ more vulnerable members, most MCs nevertheless receive some. Additional insights into why we observe differences in earmark acquisition within each political party are lacking. Yet, earmarks represent one of several legislative opportunities available to legislators with diverse interests and constituencies. These differences in desire to pursue certain legislative activities, in concert with the inherent limits on resources available to MCs, lead to a specialization in effort.
The variation with which MCs describe their Washington behavior to their constituents suggests real differences in what MCs prioritize with their time. Numerous studies point to idiosyncratic constraints on the resources of legislators, and how these constraints force MCs into difficult trade-offs. MCs exiled from preferred committee assignments compensate by shifting their effort toward more district-oriented activities to maintain their reelection prospects (Grimmer & Powell, 2013). The size of senators’ constituencies (Lee & Oppenheimer, 1999), district proximity to Washington (Fenno, 1978) and, most simply, the underlying partisanship of a member’s district (Grimmer, Messing, & Westwood, 2012; Lazarus, 2010) all alter the balance of resources MCs dedicate to different legislative activities. We extend this work into distributive politics where MCs must decide which legislative activity to pursue.
Theory
Previous research on the legislative behavior of MCs often focuses on two areas of specialization: distributive politics and policymaking. Mayhew (1974) posited that MCs will skew their efforts toward distributive politics, as it better services the reelection goals of most incumbents. From studies of the Senate, it is clear the motivations for distributive politics are not equally strong across members. Small state senators are more likely to specialize in appropriating activity whereas senators representing large populations gravitate toward policymaking activities (Lee & Oppenheimer, 1999). How the divide between the two legislative activities falls in the House of Representatives is unclear given that House districts are relatively equally apportioned. Nevertheless, the considerable differences in constituency composition alone would suggest a variety of corresponding strategies for legislative activities and resource allocation by MCs (Fenno, 1977).
For the purposes of clarity, we define policymaking effort as MC actions within the formal legislative process to draft and enact their introduced policy proposals as well as their preferred policy proposals. Broadly speaking, such activities may include everything from attending or holding committee hearings, participating in floor activity, making speeches, and introducing or sponsoring legislation. Policymaking provides the opportunity for long-term credit claiming on passage of legislation and short-run position taking. In contrast to policy activity stands particularized appropriating, with earmarking representing the classic case. Earmarks represent the classic example of legislative credit claiming because often the earmarked appropriation is targeted to specific projects within the MC’s district. For the purposes of this article, we define earmarks as any appropriation line item that designates federal funds to a specific project at the request of a MC. Earmarks allow a strategic MC to make the claim that any particular project that received funding may not otherwise have been completed without their support. For MCs with one eye on the electoral horizon, securing material benefits for their district may be a better use of their time than seeking the long-term potential in passing public policy.
Our theory builds from the assumption that MCs face constant demands on their time in Washington from committee hearings, caucus meetings, and floor votes to staff briefings, constituent meetings, and fundraising. At an institutional level, the finite resources of MCs necessitate specialization in the form of policy committees (Krehbiel, 1992). This institutional structure incentivizes the resource constrained MC to similarly specialize and prioritize activities that offer the most benefit for level of investment. The particular legislative strategy a MC pursues in light of these constraints is based on two factors: (a) individual and district characteristics that increase the value or lower the cost of an opportunity, and (b) institutional rules and norms that affect the cost of seizing an opportunity. Although a MC may recognize the benefits in pursuing any number of opportunities, the effort necessary to capitalize on any given opportunity decreases their ability to seize other opportunities.
Grimmer and Powell (2013) have previously shown that strategic MCs reprioritize their efforts in response to changes in their opportunities. Similarly, the trade-offs imposed by the individualized costs and benefits associated with different opportunities should lead to diverse portfolios of legislating activities among legislators. To test this theory we examine two common, but very different activities: the pursuit of particularistic benefits through earmarks and the pursuit of policymaking.
Certain members of the House have an institutional advantage in securing earmarks. Among these groups, members of the Appropriations Committee would seem to benefit the most. Their institutional position gives them insider access to the earmarking process. With lower barriers to inserting and keeping earmarked projects in spending bills relative to other MCs, appropriators should gain a larger share of the earmarking pie.
But, if Appropriations Committee members systematically secure a greater share of earmarks than their colleagues, then their effort dedicated to other activities may suffer as a result. Given the long-observed institutional advantage that a seat on the Appropriations Committee offers for securing earmarks, one would expect strategic legislators in such a situation to invest less effort toward other opportunities than their colleagues for whom the pursuit of earmarks is more costly. With a greater return on investment from earmarking, members of the Appropriations Committee should have less incentive to expend effort in the policy arena where the playing field is substantially flatter.
Measuring Pork Barreling and Policymaking
To test the relationship between earmarks and policymaking effort, we take advantage of a new earmark database made publicly available by the OMB (Clemens, Crespin, & Finocchiaro, 2015; Engstrom & Vanberg, 2010). 2 The OMB dataset offers a more complete picture of earmarking than past datasets, which have used the zip code of earmark recipients listed in appropriations bills to identify the congressional sponsor (Balla et al., 2002; Lee, 2005).
Our dataset includes all confirmed earmarks in the final budget identified by the OMB for the fiscal year (FY) 2010 budget during the 111th Congress. 3 The dataset specifically identifies the location of the earmark recipient and whether a legislator sponsored an earmark alone or with others. We weight the total amount of earmark dollars a legislator successfully sponsored by the number of sponsors for each earmark as the dependent variable (DV). If a MC is the sole sponsor of an earmark then they are attributed with the full value of the earmark, but when there are multiple sponsors that earmark’s dollar value is split evenly. Weighting the earmark dollars provides a better approximation of each MCs contribution toward obtaining an earmark. Consequently, our measurement of pork barreling activity substantially improves the attribution of earmark dollars to legislators than previous measures.
To measure the effort MCs exert in the policy process, we utilize a connectedness score calculated by J. H. Fowler (2006a) from bill cosponsorship data. Each legislator’s connectedness score is derived from the network of cosponsors their bills receive along with their own sponsorship patterns of bills introduced by other members (J. H. Fowler, 2006a, 2006b). Weighted to take into account MCs differing base rates of cosponsorship activity, the score reveals which members are most closely linked to the rest of Congress. 4
Although cosponsoring a bill is costless to a MC, the literature suggests that assembling bill sponsors is quite strategic. Bill cosponsorships serve as a signal to other members about a policy proposal’s viability (Kessler & Krehbiel, 1996). In this signaling game, a diverse group of cosponsors demonstrate to institutional gatekeepers that the legislative coalition in support of the legislation is broad enough to pass the bill. These relatively weak ties between legislators, nevertheless, increase a bill’s likelihood of success (Kirkland, 2011). The value of the connectedness measure comes from how it discounts costless position taking behavior relative to more strategic bill sponsoring actions that increase a policy’s chance of success. The more frequently a MC is able to attract a diverse coalition of cosponsors to their legislation, the higher their connectedness scores will be. Most importantly, these signals matter: A bill with an ideologically diverse set of cosponsors is more likely to pass through legislative hurdles (Harward & Moffett, 2010).
Each Congress has a unique set of connectedness scores, allowing for variation in MCs level of connectedness as their majority status and committee assignments change. Our use of legislative connectedness in lieu of explicit measures of policy success in Congress arises out of concerns that policy success is endogenous to majority party agenda setting, which is unlikely to mirror actual investments in policymaking of both parties. Minority and majority party members cosponsor at similar rates (Harbridge, 2015), and as a result they achieve comparable connectedness scores, even if the minority members rarely succeed in passing legislation. An added value of the connectedness measure comes from how it discounts costless position taking behavior relative to more strategic bill sponsoring actions that increase a policy’s chance of success. Cosponsoring bills with a large number of other cosponsors does not receive the same weight in the connectedness calculations as bills with fewer cosponsors.
Using connectedness as a measure of engagement in policy activity holds face validity. Table 1 shows the list of five most and least connected members of each party in the 111th Congress. Overwhelmingly, the most connected MCs are highly visible members of the chamber, with reputations for engagement in policy (J. H. Fowler, 2006a, 2006b).
The Five Most and Least Connected Members of Each Party in 2009-2010 (111th Congress).
Note. The table excludes top chamber leaders who, by refraining from cosponsoring legislation, have low connectedness scores that belie their obvious participation in policymaking. We account for these disparities in our empirical models with a leadership dummy variable.
Unlike measures of policymaking contingent on legislative advancement of one kind or another, members’ connectedness scores encapsulate the policy effort prior to roll call votes. Consequently, connectedness scores for MCs are correlated across congressional sessions when they are in the majority and in the minority. The correlation between MC connectedness in the 111th and 112th Congress is .53 whereas the correlation between Volden and Wiseman’s (2014) Legislative Effectiveness Score (derived from bills advancing in the legislative process) is –.05. Although seniority and majority party status are factors out of control of an individual legislator and thus endogenous to a congressional session, connectedness captures the specific interpersonal ties created in the actual process of cultivating support for legislation. Indeed, more highly connected MCs receive greater support for their bills on the floor (J. H. Fowler, 2006a).
As a result, connectedness is a better measure of policy effort than direct measures of legislative success, because it utilizes behaviors that precede the agenda setting power of the majority party, or gridlock due to divided government and political polarization. Legislators can express sincere policy preferences regardless of whether the majority party will allow a vote on the measure.
Controls
Previous work finds numerous other factors also influence successful earmark procurement, which we include as control variables: seniority, chamber majority, MC gender, chamber leadership, committee leaders, electorally competitive districts, district demographics, and member ideology. Existing research shows that senior members tend to be more successful at securing benefits for their constituents (Engstrom & Vanberg, 2010), which we account for by including years of service. We include dummy variables to indicate party membership, gender, and party leadership. The party leadership dummy variable takes a value of 1 if the MC is the Speaker of the House, majority, minority leaders, or party whips. We similarly indicate committee chairs or ranking member status with separate dummy variables.
To account for district demographics, we include the percent of non-white constituents by congressional district. Evidence suggests that small state legislators may be particularly focused on acquiring benefits such as earmarks (Lee & Oppenheimer, 1999). To account for this, we control for single-district state representatives: Alaska, Delaware, Montana, North Dakota, South Dakota, Vermont, and Wyoming. To measure ideology, we include the first dimension of DW-NOMINATE as a measure of ideology (Poole & Rosenthal, 1997). We also include regional fixed effects to control for any geographic variation in earmark procurement. To address influence senators may have in boosting the earmarking success of House members, we include a dummy variable if a MC’s district is in a state where the senator sits on the Senate Appropriations Committee.
Finally, past work has suggested the importance in taking care of marginal members of the majority party coalition (Cox & McCubbins, 1993; Engstrom & Vanberg, 2010; Jenkins & Monroe, 2012). This can also be viewed as the majority party ensuring their continued status as the ruling party through providing earmarks for their vulnerable members. We account for this possibility by including the last general election vote share.
Method
According to the OMB records for budget in FY 2010, 40 of the 435 members of the House of Representatives received no earmarks. This abundance of abstainers, in addition to the nonnegative, not normally distributed count data, as shown in Figure 1, violate the assumptions of ordinary least squares (OLS) specifications. To address these concerns, we estimate zero-inflated negative binomial models to test our hypotheses (Greene, 1994). The zero-inflated negative binomial regression is a two stage model, which first predicts which MCs receive zero earmarks and then predicts why some MCs are more successful than others at acquiring earmarks, conditional on receiving at least one. The negative binomial technique relies on an underlying gamma distribution to address over-dispersion in the data, which in this case is the tendency for MCs to receive more earmarks contingent on receiving at least one. The dispersion factor in the models is consistently significantly greater than zero, rejecting the null hypothesis of a Poisson model that assumes independence in successful earmark acquisition. In addition, we report results of Vuong test comparing the zero-inflated negative binomial model with a standard negative binomial model. The z value of 23.44, reported at the bottom of Table 2, is both positive and highly significant, which indicates an overpopulation of zero observations making the zero-inflated model a better fit than the standard negative binomial. For all estimations, we cluster standard errors at the state-level to account for interdistrict spillover of earmarking and policy efforts as well senators’ influence on the earmarking process. We drop one observation, Illinois’ fifth congressional district, which was vacant for much of this time period. We end up with an N of 434, including 394 legislators with at least one earmark. 5
Predicting Weighted Earmark Procurement.
Note. A Vuong test of the Zero-Inflated Negative Binomial versus Negative Binomial (z = 23.44; p < .0001) is positive and significant indicating that the zero-inflation model better fits the data creation process than the normal negative binomial. Errors clustered on state in parentheses. Control variables include the following: Population, Percent non-White, Percent Urban, Committee Chair and Ranking member Dummy, and Female.
DV= Dependent Variable.
p < .1. **p < .05. ***p < .01.
Distribution of weighted earmark dollars.
Results
The second stage of the zero-inflated negative binomial regressions presented in Table 2 confirm our primary hypothesis that as a MC increases their policymaking effort, their amount of earmarks decreases. 6 Across all the estimated equations, connectedness significantly and negatively predicts procured earmark dollars. Controlling for individual and institutional factors do not substantively change the results. Out of concern that the scale of changes in connectedness may be difficult to interpret, we also replicate the estimation in the Supplemental Appendix using a rank ordering of connectedness scores as the dependent variable. Again, the results are striking. The harder a MC works to cultivate success for their policy proposals, the lower their haul of earmark dollars. In another robustness check, we replicate the results after omitting the top 5% of earmark earners from the estimate, who may be driving the relationship. The significant result (see Supplemental Appendix) confirms that the trade-off between policy activity and pork barreling exists beyond just the tail ends of the distribution.
To better appreciate the magnitude of the impact and aid in the interpretation of the negative binomial coefficients, we present the marginal effects of connectedness in Figure 2. The impact of connectedness on a legislator’s share of earmarks is substantially greater than two common explanations of earmarking: electoral vote share and seniority. Moving from the 25th percentile to 75th percentile in connectedness decreases a MC’s predicted share of weighted earmark dollars by US$1.7 million. From the minimum to maximum level of connectedness, that difference grows to US$12 million. Table 2 also confirms our second hypothesis, that members of the Appropriations Committee are much more successful at securing earmarks than other MCs. Unsurprisingly, membership to the Appropriations Committee lowers the barrier for securing earmarks for a MC’s district. Moreover, MCs from single-district states are more successful with earmark acquisition, capitalizing on sole credit claiming ownership of all a state’s earmarks.
Marginal effects of seniority, connectedness, and vote share.
Finally, Table 2 demonstrates how past theories of earmark acquisition may be less important after accounting for legislative connectedness. Neither member ideology nor party membership is statistically significant once accounting for connectedness. Members with a lower vote share in the prior election do in fact receive more earmarks. These relationships offer additional evidence of the universalism of pork barreling during the era of earmarks. Finally, seniority is insignificant. Senior members often hold safe seats and greater institutional leverage so they have fewer incentives to pursue earmarks.
We compare the predictive power of connectedness with measures reliant on legislative success in Table 3. The alternative measures uniformly predict inverse relationships with earmarks, but only the effect of number of bills introduced by an MC reaches significance. Bill passage on the floor and enactment are heavily reliant on the acquiescence of the majority party leadership. Like bill cosponsorships, partisan gatekeepers cannot prevent a policy-oriented MC from introducing legislation. Consequently, connectedness and bills introduced better reflect the effort of policymaking than the successes of that effort, like passage thresholds. In comparison with the US$1.7 million decrease in weighted earmark dollars when moving from the 25th to 75th percentile, a similar shift in bills introduced decreases a MC’s share by US$868, 000. Column 4 of Table 3 replaces connectedness with the number of floor speeches given by a MC and negatively predicts earmark totals. The convergent predictions of the different policy activity variables further confirm the fundamental trade-off between pork barreling and policymaking.
Predicting Earmark Procurement Using Alternate Measures of Policy Activity.
Note. Standard errors clustered on state in parentheses. Control variables include the following: Population, Percent non-White, Percent Urban, Committee Chair and Ranking member Dummy, and Female.
p < .1. **p < .05. ***p < .01.
Trade-offs for Appropriators
To test whether members of the Appropriations Committee (Hypothesis 3) exert systematically less effort in policymaking than other MCs, we regress Appropriations Committee membership on connectedness scores from the 93rd Congress (1973) to the 111th (2011). This allows us to examine more than 30 years worth of committee members and their effort in policymaking. We include the same control variables as before, including party and committee leadership, seniority, and ideology. Table 4 presents the OLS regression on MCs’ connectedness.
Predicting MC Connectedness Scores Over Time.
Note. Standard errors clustered on member in parentheses. MC = Member of Congress.
p < .1. **p < .05. ***p < .01.
Table 4 shows clear and consistent results: Members of the Appropriations Committee have significantly lower levels of connectedness compared with other MCs. This stands in contrast to more senior members, who demonstrate higher levels of connectedness. The results provide compelling support for our theory that trade-offs between policymaking and appropriating extend well beyond the single session of verifiable earmark data we employ. Adding member fixed effects to the equation in column 3 of Table 4 further shows that the inverse relationship between Appropriations Committee membership and connectedness is not the product of selection of type of MC onto the committee. Joining the Appropriation Committee results in a significant decrease in a MC’s level of connectedness.
In column 4 of Table 4, we replace connectedness with the number of floor speeches a member gives as an alternative measure of policymaking effort. The model reproduces the key inverse relationship between policymaking activity and Appropriations Committee membership. Appropriations members on average give 23 fewer floor speeches per Congress than their peers. 7 Since at least the early 1970s, when our connectedness measure is first calculable, our results demonstrate that MCs have specialized in response to resource constraints.
Conclusion
The inverse relationship between policymaking effort and pork barrel success highlights the trade-offs MCs must make in dividing their attention, time, and resources. Although opportunities to burnish one’s reputation with constituents abound in Congress, realizing them requires costly actions from MCs. The negative correlation between connectedness and earmarks is substantively important with effects equal to or greater than many previously suggested explanations of pork barreling. Poorly connected legislators in the realm of policy receive nearly US$12 million more than their most connected colleagues.
We argue that the act of securing earmarks as a credit claiming strategy is an effortful action that legislators do not pursue with equal vigor. Perhaps, more interesting are the implications of treating earmarks as one of many activities legislators must choose between while confronting resource constraints. For instance, how did MCs dedicated to earmarking adapt to the 2011 ban after Republicans took control of the House? Some directed their efforts to “letter-marking,” requesting specific project funding directly from agencies, but recent work suggests MCs do not enjoy the influence over spending they previous possessed with earmarking (Mills, Kalaf-Hughes, & MacDonald, 2016). Given the increased inefficiencies for seeking particularistic projects, our theory of constrained credit claiming would predict that the ban pushed MCs to pursue alternative legislative activities, such as oversight activities or policymaking.
An important contribution of this study is the utilization of better data to measure earmarks. Our earmark data from the OMB provide not only the valid total amount of money in earmarks a MC steers toward their district, but also how many other members also cosponsored the earmark. From that network of earmark sponsorships, we offer a more precise measure of earmarking activity that takes into account multiple cosponsors. This improves on earlier earmark activity measurements that assume any earmark steered to a recipient with a zip code in a Congressional district is the exclusive product of that member’s efforts. In fact, most earmarks are cosponsored by multiple members across both chambers and frequently go to recipients outside their districts. Our study is the first to adopt J. H. Fowler’s (2006a) cosponsor connectedness statistic. The cosponsorship activities that boost connectedness, specifically the solicitation of diverse bill sponsors to signal broad support for legislation, we argue reflects purposive policymaking effort. The fact that connectedness negatively predicts earmark totals indicates differential effort expended by legislators between these different domains of activity. Recognizing the resource constraints of MCs forces a more realistic view of how legislators behave and the balance they must strike among the many opportunities available to them.
Taken as a whole, the findings clarify how scholars may think about the place earmarks and other particularistic politics take among other congressional activities. Earmarks frequently are treated in the literature as credit claiming rents taken by reelection-seeking legislators, without fully appreciating the opportunity costs incurred in their pursuit. Although recognizing the particularistic tendencies of legislators is undoubtedly important, contextualizing those actions in a broader theory provides a deeper understanding of the motivations of MCs. Like Fenno (1978), we theorize that earmarks and policymaking are part of a broad array of legislative work oriented toward appealing to member’s constituents. A simple economic framework highlights an important implication for understanding earmarking behavior: In a world without limitless resources, legislators must prioritize the pursuit of particularistic policy at the expense of other opportunities. Accounting for the effort offers a new and powerful explanation for why some legislators are more likely to pursue pork barrel projects than others.
Footnotes
Acknowledgements
The authors would like to thank James Fowler, Gary Jacobson, and Alan Wiseman, and participants of the University of California San Diego American Politics Workshop and Human Nature Group for early and insightful comments on this article.
Authors’ Note
An earlier version of this article was presented at the 2014 Midwest Political Science Association Meeting in Chicago, Illinois. The authors are listed in alphabetical order.
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) received no financial support for the research, authorship, and/or publication of this article.
