Campaign Contributions and Governmental Financial Management

In the paper, Campaign Contributions and Governmental Financial Management: Evidence from State Bond Pricing, which was recently made publicly available on SSRN, I study campaign-finance agency costs related to pricing in the $2.9 trillion state and local government bond market. By selecting a contributing underwriter directly, the government could incur significant costs with respect to government bond underpricing. There is no comparable impact when an underwriter is chosen through an auction. Through the use of a control function approach, I show that the decision to select a contributing underwriter is endogenous to first-day returns. When underpricing is expected, the government’s propensity to choose a contributing underwriter decreases as expected underpricing increases. This evidence supports the idea that there are significant agency costs associated with campaign contributions and the evidence remains robust after a battery of checks.

This paper’s results lend support to the political agency cost model. Consistent with the common assertion that the election is the primary disciplining mechanism for political executives in a political agency cost model (Besley, 2006), the likelihood that a contributing underwriter is chosen is decreasing in the closeness to the next election. Consistent with the idea that laws can discipline politicians directly and through taxpayer monitoring, the likelihood that the government does not choose an auction to select an underwriter is decreasing in the quality of conflict-of-interest laws and freedom-of-information laws; the likelihood that the government chooses a contributing underwriter is decreasing in the quality of freedom-of-information laws.

This paper’s results show, through the link between expected underpricing and underwriter choice, the important role of financial sophistication and suggest that, for government financial executives, effective financial education and competitive wages could improve taxpayer welfare. Moreover, this paper’s results lend support to the campaign-finance channel of political agency costs. The governor’s margin of victory, a proxy for other politician-agency-costs, has no effect on bond underpricing in a specification that accounts for campaign contributions; the underpricing effect is largely through the selection of a contributing underwriter. Furthermore, agency costs are present when contributions are made to the agent most responsible for contracting with the underwriter. Treasurers are the political executives most responsible for state funding. When treasurer campaign contributions are acknowledged, the results are robust.

Consider the quantitative importance of this paper’s results. Bonds are typically issued in a series. The median amount for a given serial issue is $200 million; the average amount is $293 million. Based on these figures and an average naive underpricing effect of approximately 4%, states are leaving $8 million on the table (per issue) at the median and $11.72 million per issue on average based on the naïve underpricing effect. The average two-year cumulative contribution amount in the sample is $3,026.80. For the contributing underwriters that are chosen, the average is $7,314.77; the median is $1,650.

Based on these figures, the returns to campaign contributions seem “excessive.”

No easy policy solution exists for the political agency problem presented in this paper. Any employee of an underwriter (or an associated bank holding company) has the right to contribute to his or her candidate of choice. Moreover, campaign contributions could merely serve as a proxy for side payments to political executives. But as with the case of construction firms (Rose-Ackerman, 1975), tax-paying citizens can incur substantial costs when financial service firms become noncompetitive in seeking government business. Paradoxically, this paper’s results suggest that underwriters could be competing on the dimension of campaign contributions; bond prices increase with the total amount of underwriter campaign contributions received by the governor.

A commonly proposed solution for the agency costs of security issuance requires that states use the competitive method for financing. This option has problems of its own; there could be reasons for using the negotiated method that are unrelated to agency costs. In addition to what states can do, regulators should consider applying Rule G-37 to all underwriter employees and consultants.

The full paper is available for download here.