Research note. Formal independence makes agencies more prone to politicization

Research note by Laurenz Ennser-Jedenastik

During the past decades, many public tasks have been outsourced from the central bureaucracy to agencies that operate at arm’s length from the government. This shift has been especially pronounced in the field of regulation. Formally independent agencies now regulate our energy, telecom, and financial markets, oversee the safety of drugs and food products, and enforce environmental standards.

To strengthen the credibility of these agencies, governments often award them with high levels of formal independence, where agency officials are not bound by directives of their political superiors. For instance, senior agency staff are hired on fixed-term non-renewable contracts, enjoy high levels of job protection, are forbidden to hold positions in government simultaneously, and are given organizational and financial autonomy within the agency. All these measures boost the formal autonomy of an agency from the political sphere and thus strengthen its credibility vis-à-vis stakeholders in the agency’s policy domain (e.g. businesses and consumers). Earlier research has, in fact, produced quantitative measures of formal agency independence (see here or here).

Equipped with such measures I set out to explore whether formal agency independence has any impact on the politicization of senior-level agency appointments. The intuition behind this assumption is as follows: When politicians make agencies more formally independent, they cede some degree of control over a policy domain. Even though they can always revert to enforcing their preferences through the legislative process, the nitty-gritty of policy implementation remains for bureaucrats to decide who are not under the direct command of elected officials. Since the public nevertheless holds government accountable for an agency’s performance, there is an incentive for politicians to reclaim some of the control over the agency’s dealings through informal means such as politicization.

In a recent article I test this assumption by examining whether more independent agencies are more likely to have co-partisans of the government appointed to senior positions. The study combines existing measures of formal agency independence with a newly compiled dataset on about 700 top-level appointees to over 100 regulatory agencies in 16 West European countries between 1996 and 2013. By combing through thousands of documents (e. g. annual reports, government documents, agency press releases, election lists, media reports), I collected biographical information on all appointees. I then coded whether an appointee had any link to a government party during his or her earlier professional life (e.g. holding elected or party office, working on the personal staff of politicians or parties, standing as a candidate, or simply being member of a party). This information allows me to evaluate the link between the formal independence of an agency and the politicization of its senior management. The figure below displays this relationship.


Figure 1. Share of government-affiliated appointees by agency independence


The proportion of government partisan increases from 14 % in the agencies with the lowest levels of independence to 35 % in the most independent ones. This evidence clearly suggests that politicians respond to greater levels of formal independence with higher degrees of politicization.

Of course, many other variables could be related to both, formal independence and politicization. In multivariate models I control for an agency’s resources (i.e. staff), its age, top-tier and second-tier appointments, veto players, the strength of the rule of law, the year of the appointment, as well as variation between countries, policy domains, and agencies. As the graph below shows, the relationship between formal independence and politicization still holds.


Figure 2. Predicted probability of politicized appointment by formal independence. Note: Random effects at country and agency level (left-hand panel), as well as domain and agency level (right-hand graph).


Across the empirical range of the agency independence index (indicated by the grey bars and extending from 0 to 87), the probability of a government co-partisan being appointed increases from 3 % to over 25 %. The effect of agency independence on politicization is thus not only statistically significant but also substantively important.

What are we to make of these results? First, it appears that formal independence sometimes backfires by providing incentives for politicization. It may thus undermine the very credibility it was meant to strengthen in the first place. Second, formal independence is a very imperfect measure of actual independence. This is also what other research has found. Third, politicians and political parties are responding strategically to trends of fragmentation and ‘agencification’ of the public sector. This should be some comfort to those scholars who have long feared that such developments undercut electoral mandates and democratic accountability.



The research presented here was conducted within the project ‘Party Government, Patronage, and the Regulatory State’, funded by the Austrian Science Fund (FWF), grant no. J 3409-G11.


Laurenz Ennser-JedenastikLaurenz Ennser-Jedenastik is an Assistant Professor at the Department of Government, University of Vienna. His research focuses on political parties, electoral behavior, political and administrative elites, and state institutions. He has published in journals such as Comparative Political Studies, European Journal of Political Research, Political Science Research and Methods, Governance, West European Politics, Party Politics, and Political Studies. Links: web, twitter.




One thought on “Research note. Formal independence makes agencies more prone to politicization