Elected Versus Appointed Policymakers

Featured in print Digest

How cities pick their treasurers - whether by elections or through appointments - can have an impact on their cost of borrowing. In California, cities that appoint treasurers spend 13 to 23 percent less in borrowing costs than comparable cities with elected treasurers, according to Elected Versus Appointed Policymakers: Evidence from City Treasurers (NBER Working Paper No. 15643). Were all California cities with elected treasurers to replace them with appointed ones, they could save more than $20 million collectively, author Alexander Whalley estimates.

Were all California cities with elected treasurers to replace them with appointed ones, they could save more than $20 million collectively.

Previous studies of whether elected officials or bureaucrats do a better job at controlling borrowing costs have produced mixed results. In fact, the most compelling empirical research suggests that elected electricity regulators choose lower prices than appointed regulators, and that elected judges are more encouraging toward employment-discrimination lawsuits than appointed judges. But using real-world data to prove a causal relationship is difficult, because so many other factors beyond the method of selecting public officials may play a role.

In this study's sample of 203 California cities between 1995 and 2006, for example, it is relatively easy to see that cities with elected treasurers paid 15 percent more to borrow money than cities with appointed treasurers. But the cities with elected treasurers also had more debt on average, were more likely to have directly elected mayors and clerks, and had lower per capita income and a less educated population - all of which could also influence borrowing costs. Controlling for those factors, cities with appointed treasurers paid 13 percent less to borrow than cities with elected treasurers.

To get at the issue of causality, this study further refines the sample to examine the 31 cities that held a referendum during the period to replace an elected treasurer with an appointed one. Ten cities approved such a change. Whalley argues that cities where the referendums either succeeded or failed by a very narrow margin were likely to be quite similar, on average. From that core sample, he finds that cities spend 23 percent less to borrow money if they have an appointed treasurer.

City treasurers can affect a city's borrowing costs in two ways: by issuing new debt and by refinancing existing debt. This study finds little difference between appointed and elected treasurers on the costs of issuing new debt. But on refinancing existing debt - an activity that requires significant skill and expertise - the appointed city treasurers were much better at getting lower interest rates. One reason may be that appointed treasurers often have higher levels of education (often an MBA or MPP degree) than elected treasurers do.

The findings suggest that in some cases, there may be benefits - such as reduced borrowing costs in this case - associated with assigning technical policy-making tasks to appointed officials with specialized expertise. "The results of this study also have broader implications for the organization of public good provision," Whalley concludes. "Efforts to improve governance in developing countries may well be enhanced by emulating the division of policymaking tasks in advanced democracies."

-- Laurent Belsie