Public policy making has long been characterized by students of American politics as a slow, incremental process. Take the 2010 health care reform law for instance. Politicians have attempted to enact some sort of systemic health care reform since the Eisenhower administration, yet no such reform occurred in 1994 when the Clinton administration made it’s push for such legislation, and had not in the past. Why did the health care reform law pass this time but not in the past? The answer is not so clear. The guiding work in this area was done by Professors Frank Baumgartner and Brian Jones in their book: Agendas and Instability in American Politics, where they introduce the concept of punctuated equilibrium theory from evolutionary biology to the study of politics. The basic idea behind punctuated equilibrium theory is that change happens slowly and occurs in bursts of creativity after periods of stasis.
What has become more clear over time however, is that stasis in politics is not the same as in the natural sciences. Policies that remain untouched over longer periods of time are harder to change because of interests that develop around them that make change harder. But for some policy stasis does not only create entrenched interests but can weaken policies. When a policy becomes weaker during a period of stasis or lack of change it has ‘drifted’ away from it’s original purpose. So called policy drift is rampant, as gridlock and polarization have weakened our governing institutions ability to act outright to solve policy problems.
My area of interest for the past few months has been the minimum wage, a unique case of public policy in it’s own right. To begin with, the minimum wage naturally drifts because as time passes, inflation causes the minimum wage to lose it’s value. While not an issue for the minimum wage, it is usually quite difficult to quantify policy drift. In an area like financial regulation, it would be incredibly hard to come up with an index to measure drift. On the minimum wage, the metric I constructed took this form:
The ‘drift rate’ is calculated by taking the first difference between the real value of the minimum wage in the current year and the previous maximum real value. A positive drift rate indicates a positive deviation from the previous maximum value of the minimum wage.
This means that an increase in the minimum wage beyond what it had been in the past (in inflation based dollars) would mean that the wage had drifted positively, whereas an increase that does not return the wage to it’s previous real value would result in a negative drift rate. In graphical form, it looks something like this:
This graph has two kinds of drift rates, one with just the real value of the minimum wage in 2010 dollars and the other adjusting the rate for the percentage of workers that the law will eventually cover (this occurred over time as more and more workers were legally eligible to receive the wage). With this as my dependent variable of interest, using time series statistical methods, I can further understand what variables affect the rate at which the minimum wage drifts in periods of stasis and clarify what conditions will improve the way the wage drifts.
The models that I created to study this phenomenon include a number of statistically significant variables but also have high R-squared values (from .75 to .81), meaning that the model explains a great deal of the variation over time of the drift rate. After making sure that my data follow all the assumptions necessary I feel that I can make reasonable conclusions from my work here. To be clear, my findings are not causal, but do suggest a number of important relationships. Many of the relationships I used do not tell the simplest of stories. For example, I find that while having Democratic control of Congress and the White House does help to increase the drift rate, macroeconomic variables and interest group power have much more predictive power. These results still must be interpreted cautiously. For example, one of the most universally statistically significant variables I use is the change in rate of the growth of corporate profits. As this has grown over time, it has resulted in a more negative drift rate on average. It is unclear what the mechanism that causes this relationship. Is it because the National Restaurant Association has poured so much money into lobbying efforts, or because these groups, conditional on macroeconomic performance, affect what items show up on the agenda? This is the important question going forward, but what I have found is that the state of the poor, unemployment, unions, and corporations, conditional on politics, matter in policy change on the minimum wage.