Quantitative Policy Alignment Analysis

Karl Muth argues that the analysis of policy alignment – intranationally and internationally – can be examined using quantitative tools.

The historical approach to policy alignment analysis came from comparative law and the study of political economy. This has contributed substantially to the literature on policy alignment around topics like free trade zones, military cooperation, and shipping lanes. However, the ability to do proper comparison-of-comparison meta-analysis (difference-in-differences analysis) in some quantitatively-robust context has been notoriously weak.

In a new law-and-economics paper, I explore (with my brilliant co-author and friend, Katie DeVelvis, a Lecturer in Statistics at Northwestern University) whether similarities between participation in optional schemes can be a variable used in comparative policy analysis. For instance, suppose two counties in England have the option but not the obligation (as they often do) to honour one another’s licences (for instance, a simple business licence). In some cases, there are reciprocities that develop. In some cases, the recognition is unidirectional. In some cases, there is no recognition at all. In essence, the degree of recognition of the counterpart’s policy decision (to issue a licence to a given business) can be resolved numerically as 1.0, 0.5, or 0.

Suppose now that we examine other characteristics that are quantifiable, perhaps council tax regimes or the level of fees charged for rendering a given state service. We would hypothesise that, all else equal, the pairs of counties with similar fees or taxes would be more likely to have reciprocity relationships. In the American context, we’ve embarked on precisely this analysis. We have built clusters of American states that are geographically similar and researched their rates of taxation for various activities and then also looked at their recognition or non-recognition of similar licences issued by a neighbouring state’s government.

Of course, some states are more xenophobic than others (Ohio, for instance, rarely recognises any administrative decision or licence granted by another state), but for the most part states have a mix of either reciprocal, unidirectional, or non-recognition relationships with the licences issued by other states in the region. This provides fertile ground for this type of comparative analysis.

Much as George Kingsley Zipf and his contemporaries brought quantitative, rigorous analysis to linguistics three generations ago, I believe the analysis of policy alignment – intranationally and internationally – can be more carefully examined using quantitative tools. By comparing wholly elective alignments between jurisdictions (aside from vehicle number plates and a handful of other examples, there is no requirement that one jurisdiction respect its’ neighbouring jurisdictions’ decisions as to decisions and licences issued) with policies that are set within a given jurisdiction (county-level or state-level taxation, for instance), we see the degree to which policies are aligned on a level deeper than observing mere similarity. I am hopeful this paper, and subsequent scholarship in this area, will build tools that allow jurisdictions to be compared not only in terms of subjective observations as to their political alignment, but through tools that allow both nuanced analysis of policy alignment in the present and political momentum aimed at the same point on the horizon over time.

The paper, expected to be published in 2015, is K. Muth and K. DeVelvis, Testing Policy Alignment: A Quantitative Methodology (Forthcoming 2015).