I began working on “Political Parties and Macroeconomic Policy” in 1974, delivered the paper at the August 1975 meetings of the World Econometric Society in Toronto and the APSA in San Francisco, and shortly thereafter submitted it to the Review, where it languished for more than a year before it was accepted (just barely, I sensed) after a change of editor. The article's title is somewhat misleading because the evidence pertained to the influence of government partisanship on macroeconomic outcomes.

Analysis of how partisan forces affect intervening policies came afterward in a series of my own works (Hibbs 1986, 1987, 1994; Hibbs and Dennis 1988) and in important research by many others. The closing sentences of the 1977 article sum up what motivated my research: “Macroeconomic outcomes…are influenced to a significant extent by long- and short-term political choices. The real winners of elections are perhaps best determined by examining the policy consequences of partisan change rather than by simply tallying the votes.”

My orientation departed from those prevailing at the time. In the 1960s and early 1970s, economists viewed policy analysis mainly in terms of benign government planners aiming to maximize “social welfare” functions subject to constraints. During the same period, political scientists regarded voting and elections as largely sui generis—as autonomous end points of research. The agenda-setting framework for the political economy of electoral democracies was Downs's monumental 1957 treatise An Economic Theory of Democracy, which assumed purely office-motivated policy behavior. I felt that those orientations mistakenly neglected conflicts of economic interest among voters and associated cleavages in economic objectives among parties that are fundamental forces shaping macroeconomic policies and outcomes.

During the course of my research I learned of an unpublished 1974 paper by Tufte on electoral cycles in economic policy (which formed the core of his famous 1978 book Political Control of the Economy) and of the prepublication version of Nordhaus's famous 1975 article “The Political Business Cycle”—both based on the Downsian assumption of election-driven policy. The Tufte and Nordhaus papers prompted me to test for election period signals in U.S. unemployment as a side hypothesis to my primary focus on partisan influences, and in a footnote I reported being unable to detect significant effects related to the U.S. presidential election calendar. The evidence accumulated over the subsequent 30 years in hundreds of studies of dozens of democracies by and large squares with the inferences that I drew from sparse data in my 1977 article: partisan effects on macroeconomic variables are commonplace, whereas election calendar effects are sporadic, although, as Drazen (2001) pointed out in a review article, there are signs of regular preelection movements in fiscal variables of the sort featured in Tufte's work.

The political–distributional foundations that I laid out for the partisan model—that the interests and preferences of core party constituencies predispose Left governments to pursue relatively expansive policies intended to raise growth and lower unemployment and Right governments to pursue relatively restrictive policies designed to contain inflation—did not arouse much controversy. The main object of contention was the macroeconomic foundation: an exploitable Phillips curve hinging on “nonrational” adaptive price expectations. That foundation was dealt a crippling blow in academic circles by the rational expectations revolution in macro theory that was sweeping the hearts and minds of up-and-coming economists just as my work on partisan cycles (and Nordhaus' work on electoral cycles) appeared.

About a decade later, partisan models incorporating rational expectations and derived from nominal “surprise” theories of business cycles then in vogue were proposed by Chappell and Keech (1986) and Alesina (1987). These so-called “rational partisan theory” setups rested on the hypothesis that monetary (inflation) surprises created by unanticipated election victories by parties with divergent macroeconomic objectives were the source of partisan effects on output and unemployment observed in data. However, despite the best efforts of Alesina and collaborators to make the case for partisan policy surprise models in a series of empirical studies, surprise theories of business cycles, whether or not rooted in unanticipated changes of governing parties, have been abandoned owing to the lack of supporting evidence.

The facts are that unanticipated changes in nominal aggregates (the money supply, the price level) have nothing much to do with expansions and contractions of output and employment, while Phillips curves founded on sluggish price adjustment yield remarkably good fits to the data. For those reasons traditional Phillips curves—which never died among politicians and central bankers, most of whom always perceived a dynamic policy tradeoff between inflation and unemployment—have since made a comeback among mainstream, empirically grounded macro-economists. A good example is Mankiw's 2001 article “The Inexorable and Mysterious Tradeoff Between Inflation and Unemployment.”

My own program for advancing partisan theory was spelled out in my 1994 article. The central idea is that partisan policymakers are uncertain about the structure of the macroeconomy and hence about the impact of policies on stimulation of extra output and employment as opposed to needless extra inflation. Partisan objectives are consequently adjusted dynamically in the light of experience. Franzese and Jusko (2006) review a large body of research on partisan cycles conditioned on a much broader range of structural, institutional, and strategic contexts. This rich literature takes us in the right direction and likely will progress steadily and productively.