A test of the institutionally-induced equilibrium hypothesis: on the limited fiscal impact of two celebrity governors

Arnold Schwarzenegger was born in 1947 in Austria and is a body builder and well known actor. He began lifting weights at the age of 15 and continued with great success in international body building and weight lifting events including Mr. Olympics competitions which he won 7 times. He then transitioned to movie acting beginning with Conan the Barbarian in 1982 and continuing in a number of heroic roles to which his physique contributed to his image and performances. He was elected governor of California in 2003 running as a republican in a recall election, the only California governor so elected. Because of his prominence as an action hero, the importance of California, and some of his comments while governor (as with “girlie men”) his governorship received considerable coverage during his time in office. It was his first and only elective office. He won reelection in 2006 (his first term being shorter than normal because it was the result of a recall election). He returned to acting and producing films after leaving the governorship.

Jessie Ventura was born in 1951 and is a retired professional wrestler—a sport that is highly choreographed and dramatized. His “ring name” was “the Body, Ventura” in that entertainment venue (1975–1986). After that career, he became an announcer for professional wrestling matches and subsequently host of a radio talk show in Minnesota. He was elected Mayor of Brooklyn Park Minnesota from (1991 to 1995) and elected to governor of Minnesota in 1998, running a low-budget grass roots campaign that urged voters “not to vote for politics as usual.” He won as a third-party candidate of the “reform party.” He did not run for reelection in 2003 after his term was up. He was the only third-party candidate to be elected to governorship in the past 50 years. (A number of governors have been independents, meaning not formally aligned with a political party during their term of office, but all were previously active members of one of the two major political parties, as with Crist (Republican), King (Democrat), Weiker (Republican), and Hickel (Republican).)

Both governors were entrepreneurs and entrepreneurial, but neither were students of politics or public policy. Both rose to office in unique ways that could not have been anticipated a year or two before they happened.

The exception to that rule is Nebraska, which has a unicameral legislature. An interest group explanation for differences in the lengths of state constitutions is provided in Crain and Tollison (1979).

States receive from 20 to 45% of their revenues from the federal government with an average of just over 31.5% and a median of about 33% (taxfoundation.org/maps). Much of the revenue is earmarked for particular services such as Medicaid, education, and highways. These funds are regarded to be exogenous to state budgets for the purposes of this paper, although some may affect the distribution of state spending, governors have relatively little influence on the National government’s decisions to allocate funds to programs or projects within particular states.

This literature arguably began with Grier and Tullock (1989) and Congleton (1992) who demonstrated that democracies and dictatorships had systematic patterns of economic development and public policy, and continued through Knack and Keefer (1995) and Persson et al. (2000) and continues to the present. A useful metastudy of the effects of democratic institutions was undertaken by Doucouliagos and Ulubasoglu (2008) who conclude that democracies have indirect effects on growth through policy decisions that affect education, inflation, and economic freedom.

Influential studies of the impact of leaders and leadership characteristics on public policies include, Dreher et al. (2009) who find that past careers as entrepreneurs are more important for policy reforms than is the educational attainment of national leaders, Besley et al. (2011), who find a link between economic growth and the educational attainment of leaders, Congleton and Zhang (2013) who find that the effectiveness of U. S. Presidents as economic stewards were improved by education and particular career paths, and Dal Bó et al. (2017) who find that leaders (in Sweden) tend to be better educated and better leaders than the average Swede. (That elected leaders tend to be smarter than the average voter is a conclusion that would not have surprised Aristotle who regarded classical Greek versions of representative democracy to be forms of aristocracy. Many of these regularities are thus longstanding “facts” concerning leadership within political theory, although the statistical affirmation of these relationships is new.)

The idea that institutions and their associated bargaining and oversight produce stable public policies did not, of course, end with the first-generation papers. Shepsle and Weingast (1981), for example, argue that the micro-organization of the U.S. legislature tends to stabilize coalitions within Congress and allow long term bargains to be worked out and implemented. Congleton (1982) notes that the bureaucracy’s discretion over public policy further dampens the effects of policy shifts. Such stabilizing effects are taken for granted in most papers that attribute significant effects to constitutional architecture.

This is not to say that these men had absolutely no political experience. Ventura had been elected and served as the mayor of Brooklyn Park Minnesota, a city of 70,000 residents. Schwartznegger had served as chairman of California’s Council on Physical Fitness. However, it is clear that they lack the usual connections with their state political parties and experience in statewide offices that most governors have.

Autoregressive models have previously been shown to explain most but not all of the path of national fiscal policies in the United States. See, for example, Atesoglu and Congleton (1982).

This model may seem a bit simplistic, but consider the following structural representation of median voter demand for government service G. Each voter maximizes a utility function U = u(G, X) where G is government service and X is their private good consumption. Because turnout rises with income, the median voter’s income can be approximated with average income, Y^A. The median voter’s private constraint is X = [1 − t(G, N, Y^A)]Y^A and her public constraint is c(G) = tNY^A where N is the adult state population, c(G) is the cost of public services, and t is the average tax rate. The tax rate will be a function of service level, population, and average income. Maximizing utility generates a reduced form demand for government services of the form Gj* = γ(Yj^A, Nj) for state j with population Nj, and will be approximately (Gj*/Nj) = g(Yj^A) for per capital government expenditures. Treating G as a vector of services would not change the variable(s) in the reduced form. State fixed effects account for other differences in state demographics that might affect the preferences of the median voter.

Such partisan effects are consistent with electoral competition models that include roles for political parties. In partisan models, parties create and maintain distinct policy agendas to retain their base of supporters (Duverger 1963; Alesina 1988; Grofman and Lijphart 2003). In such cases, one would expect to observe partisan effects but not office-holder effects, insofar as parties select their candidates for high office.

https://​bea.​gov/​index.​htm.

http://​slfdqs.​taxpolicycenter.​org/​pages.​cfm.

https://​www.​leg.​state.​mn.​us/​.

http://​www.​legislature.​ca.​gov/​.

http://​www.​capitolmuseum.​ca.​gov/​.

https://​ballotpedia.​org/​State_​Legislatures.

Deficit per capita are calculated as the difference between real expenditures per capita and real taxes per capita.

For overviews of how differences in democratic institutions affect public policies, see Congleton and Swedenborg (2006). For analysis of the U.S. institutions, see Buchanan and Tullock (1962) or Tsebelis (2002) for a more general analysis of how veto players affect public policy.