A brilliant series of papers have been produced recently on the topic of U.S. Supreme Court justices' ideological change over time. Authors on these papers include Andrew Martin, Kevin Quinn, Lee Epstein, and Jeff Segal, although the author groupings change by paper. Most of the papers can be found on Quinn's website. To put this very briefly, the authors suggest that - contrary to what prior research claims - Justices' ideological preferences, or ideal points, are not time-invariant, but rather change over time (or at least most of them do). The authors show that justices' revealed prefrences (i.e. voting) change over time, independent of factors that aren't justice-specific that might affect voting (e.g. changes in the type of cases the Court hears).
In "Assessing Preference Change on the U.S. Supreme Court," Martin and Quinn make a very strong case that justices' ideal points, as revealed in their voting, are not time-invariant. This has important implications for the way we study the Court. However, in the conclusion they posit a very important question that they leave to future research: What explains preference change? We basically, need a measure (exogenous to justice votes) to provide a representation of the justices' sincere ideological beliefs over time. We have time-invariant measures, such as party identification or social background, but coming up with time-varying measures presents a bit of a puzzle. The authors suggest that we can look toward factors affecting preferences offered by social psychology, stragic considerations, and the macro-political context.
I am working with two UGA graduate students, Brian Levey and Justin Moeller, on the early stages of a project that seeks to answer just this type of question. While we have certainly not "cracked the code," on this puzzle, but we are working toward a strategy for explaining change in ideology over a person's life. We hope to develop a method of explaining revealed ideological preferences that can be applied to policy makers generally. Arguments have been made that, for instance, members of congress and presidents change their attitudes over time. Explaining such change is easier said than done, to be sure.
With this goal in mind, I'd like to take a step back and address the basic question - assuming that people do change their sincere ideological preferences over time, what causes such change? Do you know someone whose ideological preferences have changed over time? If so, what do you suspect caused this change? Is it possible that we could apply some of the methods used by Neal Tate (1981) in his seminal article using social background factors (e.g. father's occupation, justice's education, etc.) to explain justices' changing ideological preferences, as revealed in voting?
Of course, there is a rich behavioral literature in political science and other disciplines on how citizens' attitudes are formed, and some literature to suggest why they change. It is possible that justices' attitudes are changed over time by the collective preferences of their social networks. Thus, a change in ones social network might lead to a change in their attitudes or ideological preferences. Other socializing events and phenomena may also prompt change in ideology.
If we consider Supreme Court justices as an example, such events or phenomena (occuring after their confirmation) might be: where they live (change in residence or significant time spent in new surroundings), changes in personal status (marriage, divorce, children, grandchildren, death of a family member), or role on the bench (more senior, elevation to CJ, etc.), among others. There could also be interactive concerns whereby changes in personal status affect justices differently based upon certain conditioning factors.
Of course, this is all just my early musings on this question, but I would be interested in hearing what other people have to say on this topic.
I'm glad that Tracy brought up the point about public mood - I was thinking about a separate post on this topic, but I believe that this is a good place to discuss this topic. Two things come to mind:
1) While some will continue to disagree about the influence of public opinion on justice voting, there is a strong literature to support this notion. However, I think that it is important to think about what public opinion influence denotes, behaviorally. First, there is the idea that justices act strategically in relation to public opinion as well as congress, and the president. It seems that this is distinguishable from the second path of influence, that justices, like all human beings, are influenced by the world around them. It seems that the latter points to changes in justices' sincere ideological beliefs while the former denotes strategic behavior and the I think that there may be some problems in combining the two as one concept. BTW, an interesting set of "reads" on the influence of public opinion on the Court can be found in the writings and public speeches of William Rehnquist.
2) Prior assessments of the influence of public opinion on SCT justices have typically focused on national public mood, often using Stimson's national measure. However, we might wonder if justices tap multiple sources of public opinion in shaping their everyday lives and work decisions. For instance, national opinion might be very relevant, but what if it's different from the public opinion in the area where they live and work (VA, DE, MD, or DC), day in and day out. Alternatively, does the mood of their "home" state matter - where they came from. I imagine that some justices keep stronger ties with their "home" than others, but this might be something to consider. Finally, this point only touches on geographic differences, but public opinion also varies by income level, education, and a host of other relevant concerns.
All of this might be something to consider when we think about the ways that public opinion might matter to justices.
Posted by: Jeff Yates | 26 January 2007 at 12:12 PM
A worthy endeavor and a necessary one. A problem and a suggestion.
One reason the Segal-Cover scores have ruled so long is that there is substantial evidence that basic ideological preferences don't change all that much over lifetimes. The evidence cited in the papers above is almost certainly, as Martin and Quinn say, the product of ideosyncratic events in individual justices's lives. I would bet that, if we could ask them and get a straight answer, the justices would all say that their basic views haven't really changed all that much. They were just reacting to what they percieve as changes in their colleagues. Some sort of comparison of shifts in overall ideological composition on the court and preference changes over time might prove fruitful.
I would also suggest that you try to track changes in preferences to changes in general public opinion. The justices are, when they lay their robes aside, regular members of the elite tribe of the citizen body. We know that their decisions track public opinion shifts at a lag. It would be interesting to see if their preferences do too.
Posted by: Tracy Lightcap | 26 January 2007 at 11:30 AM
Kirk, thanks for your comment; it allows me to develop a point on which I am afraid I wasnt very clear. - I'm not planning to use the time-invariant Tate background factors alone to assess voting preferences over time, but rather I plan to use a similar methodology (index of new life events, changes in social networks and lifestyle conditions, etc.) that occur after justices' confirimations, across the justices' tenure on the high court. So, true, we might start out with a static score upon confirmation as a baseline (e.g. Tate, Segal/Cover, etc.), but then the baseline would change as these events and phenomena occur over time. Some justices would have more change factors than others, but there would probably be changes for all justices on some dimensions (e.g. age, perhaps the # of opinions idea that Kirk mentions). I hope that this clarifies things a bit.
Posted by: Jeff Yates | 26 January 2007 at 09:13 AM
I think Jeff raises a very interesting and extremely important question here. I think most of us would acknowledge that preferences change over time, but developing a measure of those changes is extremely difficult. And, trying to explain why those preferences change is even harder. Jeff mentioned possibly using the methodology employed by Tate (1981) is his social backgrounds article. However, if we wish to explain changes in preferences, we need to examine factors/predictors which themselves change over time. Static measures, such as educational background or previous occupational experience will not lend much insight into later preference changes.
So, we are left with trying to identify events that occur during a justice's tenure that might influence a change in preferences. Perhaps our current measures of legal expertise (i.e. the number of opinions written in a particular area) might be useful. As individuals gain more expertise in specific areas, the additional accumulated knowledge might influence policy preferences.
This is just one example of an indicator that itself changes over time, and therefore might influence preference change. If we can identify other indicators with temporal properties, then perhaps we can gain some analytical leverage on this question. Not an easy task, but certainly an important one.
Posted by: Kirk Randazzo | 26 January 2007 at 08:54 AM