No, really, what is APD?

In recent posts, two of my fellow bloggers – Anna Law and Steven White – have written about the question of what APD is (and, what it isn’t). For those people who are in political science, but not too familiar with APD as a specific field of study the fact that this question is being debated might be surprising or confusing. We generally know what something like behaviorism is, or what a courts scholar does. Why is APD so complicated?

And yet, it is.

In January, the Bedrosian Center and The Society for Political Methodology organized a two day conference on the relationship between causal inference and American Political Development. The goal of the conference was to see whether there were ways either ‘side’ could learn from the other. However, one of the ongoing questions that was raised by nearly all participants – both those who are not self-identified APD scholars, and those who are – was the ongoing confusion over what APD actually means. And, even after two days of healthy debate, I don’t think the question was settled.

The lack of clarity has not been helped by the existing definitions of the term. The most famous one has been posited by Orren and Skowronek – two of the founders of the APD approach – who argued that APD is the study of “durable shifts in governing authority” – where governing authority concerns “the exercise of control over persons or things that is designated and enforceable by the state,” a shift is “a change in the locus or direction of control” in the distribution of authority, and durability is intended to acknowledge that history by definition means change over time.

There are some issues with this way of defining APD. First, the focus on governing authority is rather limiting. While studying the development of the federal bureaucracy over time can fit comfortably within the definition, other things are not directly related to the state but can still be studied by political scientists. For example, a major strain of my own research focuses on the role of national party organizations. Those party organizations are not part of the state, or its exercise of control over persons or things. Thus, those studies cannot be considered APD (unless we define ‘the state’ as being anything to do with politics in the US, at which point the term becomes almost meaningless).

That being said, the Orren and Skowronek perspective does actually give us some clarity as to what APD is and what it is not. While the governing authority element may be too limiting, the idea of focusing on shifts is still appealing. In particular, the focus on ‘shifts’ helps us distinguish between those studies that focus on historical change over time, and those that use historical data or case studies with other purposes in mind.

With that in mind, I would add three general points to those made by Anna and Steven:

First, we have a tendency of combining APD studies and historical institutionalism (or historical behavioralism) studies in the same category because both rely on historical data or cases in their analysis. While there’s no real harm in doing this, it is missing that crucial difference of studying ‘shifts’: APD work looks at change over time (in terms of its causes, its consequences, or both). Historical institutionalism relies on history to test more generalizable claims about the unit of analysis.

To be sure, historical institutionalists still rely on deep historical knowledge to place their analysis in context. That is, we use that knowledge to identify why a theory may be time-bound, what the appropriate control variables are, how to correctly measure our unit of analysis for that particular period of time, or what could be possible context-specific explanations for outlier cases where the theory does not work.

One example of what I would identify as a mis-categorization in this regard is Daniel Galvin’s Presidential Party Building. This book is commonly placed in the APD category, and Galvin identifies as an APD scholar. However, Galvin’s core concern in the book is being able to explain why different types of presidents act differently towards their national party organizations. Development of either the presidency, or even really national party organizations, isn’t the main concern of the book. Obviously, the most important element here is that Galvin produced a fantastic book which – if you are at all interested in presidents, parties, and particularly their relationship – you should read (Trust me, it’s great. Here’s the Amazon link.) But in terms of defining what APD is and what it is not, it is helpful to identify distinctions like these. Note also that individual scholars can produce work that fits into either category: almost all of my research projects are historical of some type, but some are APD, and some are historical institutionalist.

Second, to paraphrase a point comparative scholar and qualitative methodologist David Waldner has made, APD is not a method, it’s a research interest. That is an important difference: APD presents a general way of thinking about topics worth investigating. How did the GOP become a mostly all-white party? How did we end up with a polarized party system? These are all questions that get at development over time, and can produce studies that would fit in the APD area of research (again, assuming we drop the governing authority requirement). But the questions do not inherently provide us with a methodological approach that allows us to answer them.

In recent years, we’ve seen a more heterogeneous methodological approach to APD studies – think of Eric Schickler’s excellent Racial Realignment – but traditionally most people active in the APD area rely on qualitative case studies. Given that many of us are trying to answer questions or test theories that cannot be done quantitatively (let alone through experiments), this is often inescapable – and, on the bright side, provides those studies with considerable internal validity. However, it should be noted that a lot of the research in APD hasn’t been particularly clear about what kind of benefits and limitations come with those methodological choices. In Comparative Politics, qualitative methods has developed more and it might be helpful if Americanists who do either APD or historical institutionalism from a qualitative perspective follow their example.

Third, while I agree with Anna that there is a difference between APD and history, I am not entirely sure whether one must be a political scientist to do APD. That is, there are a number of scholars in History who do work that qualifies as APD. Brian Balogh’s A Government Out of Sight, Leah Wright Rigueur’s The Loneliness of the Black Republican, and Sam Rosenfeld’s The Polarizers are all (really good!) APD books. All three of these scholars have PhDs in History, and two of them work in History departments (Sam is now in a political science department). That might seem confusing, unless we keep point 2 in mind: APD is a research topic, and it can be handled through different methodological approaches, including those inherent in History as a field of study.

Leave a Reply

Your email address will not be published. Required fields are marked *