At the Public Law and Legal Theory Workshop on October 27, 2010, Sanford Gordon from the Wilf Family Department of Politics of New York University answered questions about his paper, Executive Control v. Bureaucratic Insulation: Evidence from Federal Contracting, which examines political control over bureaucratic institutions and the influence that politics can have on spending, using the 2007 GSA scandal as an empirical test. The paper can be found here.
What influence do elected executives have on government expenditures?
In the paper, Gordon evaluates the amount of influence that political actors can have on government expenditures by an officially neutral bureaucracy. This inquiry requires an understanding of an executive's campaign strategy and the executive’s beliefs as to where federal money is best spent. In other words, executives must decide which voting districts they strategically wish to capture by awarding government expenditures in advance of elections. There are disagreements as to whether a political party's best strategy would be to target core constituencies, marginal constituencies, or a combination of the two.
Is an executive's chosen campaign strategy furthered by autonomous government agencies? If an agency is sufficiently insulated from the desires of a politically motivated executive, the risk is small. However, bureaucrats often are appointed to their position by a political party for their strong allegiance or ability to further the political party's goals. On the other hand, the agencies may be reluctant to set aside their perceived independence and neutrality in order to benefit their elected superiors and/or funnel federal money away from otherwise deserving recipients. Additionally, the passage of the Hatch Act prohibits many federal employees from using their position to influence elections. Lastly, there is a question of how much discretion any individual bureaucrat has to award federal money, given the bureaucratic process of most agencies.
Gordon uses the 2007 General Services Administration (GSA) scandal and exposure to explore this topic using the details of executives priorities and the subsequent federal expenditures made by GSA.
After the midterm elections in 2006, the Republican party lost the majority in both the House and Senate, and many Democratic candidates also replaced incumbent Republican state governors and legislators. A few months later, a deputy from the White House gave a powerpoint presentation to the General Services Administration (GSA), which is generally responsible for assigning contracts to private vendors for supplies or building acquisition for other agencies of the government (managing $500 billion in Federal Assets1). The presentation laid out polling data from the 2006 elections to 40 GSA administrators. Based on the data, the presentation labeled voting districts as "defense" or "targets" in the 2008 election. At the conclusion of the presentation, Lurita Doan, a GSA administrator, was reported to ask "How can we help our candidates?"
The presentation and Doan's statement were subsequently leaked to the The Washington Post, which exposed the scandal on March 26, 2007. The exposure led to negative publicity for the Bush administration and the eventual resignation of Doan who was facing Hatch Act violation proceedings.
The scandal shows the executive attempting to politicize the GSA, or, in other words, using federal contracting for Republican electoral purposes. In this paper, Gordon takes advantage of the transparency offered by the scandal to evaluate the influence an executive's electoral goals have on bureaucratic allocations of federal expenditures. The executive priorities laid out in the powerpoint offer a clear sample group to evaluate the effect of attempted politicization on subsequent federal contract awards by GSA.
Data and Analysis
Gordon focused on the effect of subsequent GSA contracting in districts that were labeled "target" or "defense" in the powerpoint. GSA is comprised of two major bodies: the Federal Supply Service (FSS) and the Public Building Services (PBS). The two bodies differed in their responsiveness to the presentation. While PBS clearly responded and increased federal contracts granted to vulnerable Republican districts after the presentation, FSS's actions are scattered and largely unsusceptible to the politicization. Gordon finds that this is probably due to a variance in discretion and relative market competitiveness for the two bodies. GSA is structured so that a bureaucrat is more constrained when a contract is more heavily bid for, and it is assumed that FSS is generally more constrained by this competitive bidding.
Given that GSA is a large bureaucratic institution, it is perplexing to imagine how an executive directive could be so effective at influencing PBS contracts in such a short period of time (59 days before media exposure and subsequent decrease in vulnerable district expenditures). Gordon offers a few explanations for this, but ultimately finds that GSA increased the dollar amount for existing agreements with vendors as opposed to soliciting new ones.
The White House was able to politicize GSA for electoral advantage, and the data shows no rational anticipation by employees of the federal contracting strategy. Thus, the Bush White House engaged in active politicization, or coordination with executives, as opposed to passive politicization, where executives appoint loyalists who work to further electoral goals independently and with little coordination.
In addition, Gordon's findings suggest that measures of electoral vulnerability perceived by a political party are based on some expert or unknown criteria (not simply previous vote margins). Finally, the results show that the success of politicization is dependent on the agency's constraints and discretion.
Discussion and Comments
At the workshop, there was a lot of discussion surrounding the proper control group for the data. Should the control group be districts that were considered "safe"? Gordon indicated that the analysis is dealing primarily with the expert White House valuation of districts and the change in spending in those districts that were indicated.
The comments of Lurita Doan after the presentation were also discussed. Gordon suggested that it seems that but for those comments, the presentation would not be a Hatch Act violation, because it was informational and the effort to politicize was implicit rather than explicit.
The discussion and paper seem to suggest that there is a lack of agency insulation from politics. Instead, if we want to deter politicization, there needs to be less agency discretion--at least in the realm of government expenditures. Although the Hatch Act is intended to combat this problem, there is a question of how often it is enforced and if it is actually effective at deterring federal employees from giving their political parties advantages. Finally, it would be helpful to see the correlation between politicization and candidate success.