The Difficulty of Institutional Reform

January 3, 2012

Happy New Year!

I’m not going to comment much on Iowa/caucuses today, because it’s so far from my field of competence that my thoughts probably have no marginal value, and definitely none in reference to the value of your time. I do want to, however, point you toward a few posts that I think are definitely worth reading: Nate Silver did his final prediction round-up here, Jonathan Bernstein explained how to think about Iowa here and here, Brendan Nyhan made a key point about media coverage here, and Ezra Klein had a nice distillation of everything already mentioned here. For a bit more whimsy, you might also want to try Dave Weigel’s post on the twitter cliches of Iowa coverage, which got me laughing.

I’d also refer you to Bernstein’s election day patriotism post. That stuff always gets me. As you probably know, I’m a total romantic sucker for the pageantry and symbolism and raw logistics of the democratic process, even the messy parts. But I’m also a pretty hard-nosed realist about the fragility of open societies and the difficulty of instituting and maintaining a stable democracy. It’s absolutely a blessing for modern America that such a government has been handed down to us, and it’s a triumph for human freedom every time another nation of people successfully unshackles from authoritarianism and realizes such a system.** I don’t think there’s any question that democracy is objectively the least-worst real-world system known to man. And so while I’m a huge supporter of citizens criticizing our political leaders and government institutions and electoral processes, I also think it’s important that people remember how fortunate we are to live under such an imperfect form of government. And there’s no better time for that than election day.

This also brings me to what I want to discuss today: institutional reform. And I mean that in the widest sense possible — everything from the above-mentioned toppling of dictators in favor of democracy all the way down to changing chamber rules in Congress or the partisan primary calendars. Now, there’s always a lot of talk about institutional reform, but I can’t really remember a time when it seemed so constantly at the forefront of the political agenda. Maybe I’m just a hammer seeing a lot of nails, but I feel like it has become a secondary story to just about everything: in foreign policy, you’ve got the shakeout of the U.S. leaving Iraq, plus the Arab Spring in Egypt and elsewhere, all encased within a continuing debate over whether regime change is a legitimate or successful basis for a foreign policy of the United States. Back at home, we’ve got the ever-more-simmering debates over the filibuster in the Senate, the quadrennial rancor over the primary process for presidential nominations, and, oh yes, seemingly everyone and their radical-moderate brother trying to undo the partisan tension in Washington.

In the abstract, it seems great that there are well-organized efforts to deliver institutional change to American politics; it’s almost certainly the sign of a healthy democracy (or at least a sign against a democracy in fast decay). But it’s hard to get enthused when you read the specific reform proposals of such outfits as No Labels or Americans Elect. In fact, it’s rather depressing. Take the 12 point action plan of No Lables, for example. All of their proposals tend to fall into one of three categories: ultimately useless ( i.e. having Members of Congress not sit by party during the State of the Union address; barring Members from taking pledges; off-the-record bipartisan gatherings; joint leadership committees; 5-day congressional workweeks; nonpartisan fiscal report to Congress), plausibly good but not exactly novel (i.e. end the filibuster; presidential question time; anonymous discharge petitions), and horrible ideas and/or dangerous cures and/or unconstitutional (no negative campaigns; automatic confirmation of presidential nominees absent Senate up/down action; blocking Member pay until appropriations are done).

And those are just the proposals, forget about how on Earth you’d get them implemented. And this leads to the question at hand: why is institutional change so difficult? I would suggest that, with the broadest brush, there are three answers to this question: conflicts between normative values and strategic self-interests; the wide scope of institutional reform; and biases against change. I think that most people assign greater weight to the latter (the anti-change biases), but my sense is that the first two are much more fundamental. Let’s take a closer look:

1. Conflict between normative values and strategic self-interests. In a nutshell, the idea here is that your reform proposal might be a great idea that everybody loves in a vacuum, but it’s almost certainly going to result in political outcomes that are different than they otherwise would have been. And therefore, it’s going to make political actors consider how the outcomes differ in the short-run and weigh that against the normative long-term value of the reform. And you will not get much support for your reform from short-term outcome losers.

This isn’t rocket-science, but I don’t think you can overestimate how important it is in all situations of potential institutional change. It’s almost so obvious that it goes unspoken. After all, the implicit purpose of institutional reform is to affect how fixed preferences are translated into outcomes. If the outcomes aren’t going to change, then your proposed reform isn’t good or bad, it’s just a waste of time, like the proposal to have Members sit in a bipartisan fashion at the State of the Union Address. But when the outcomes are going to change, in many cases the normative value of the reform ceases to even be a consideration among political actors or voters. Interests dominate. (And that’s probably a good thing, but it’s definitely a consequential thing). Like Bryan said, the people of Nebraska are for silver and I’m for silver, I’ll look up the arguments later.

Similarly, trying to assess the normative value of an institutional change is really hard, because of the strategic interests of the actors. Good government altruism aside, institutional change is often proposed specifically for the purpose of adjusting particular outcomes. Again, this is so obvious it’s barely mentioned. But it has an important effect: the normative value of the reform itself is difficult to gauge. Is it being proposed for pure intentions, or to shift short-term outcomes? Likewise, the motivations of the short-term winners are going to be questioned: do they really think it’s a good idea to get rid of the filibuster and have some other state besides Iowa go first int he primary process, or will those things just help them pass their current legislation and give their preferred candidate a better chance at the nomination and their home state more say in the process?

Consider something like the voting age. It would be (and was) a significant institutional change to raise or lower it. And there might very well be strong normative arguments for letting 14 year-olds vote, or for returning the constitutional maximum-minimum voting age to 21. But there’s little chance such normative arguments would play the role of anything but just that, arguments. Much more likely is that the outcome of such a move would be a prime consideration of any supporter or opponent. How would a lower or higher voting age play out? It’s not obvious, but the general intuition is that a lower age might help liberals, while a higher age might help conservatives. And you can bet dollars to donuts that one sentence would be a better predictor of opinion than almost anything else. And the normative arguments would simply flow from there.

And thus the situation almost always becomes one in which normative values cannot be trusted: they may be genuine or they may be strategic rationalizations for underlying short-term interest winners. From the point of view of an individual political actor or voter, the inability to trust the normative information creates an incentive to rely on an evaluation of the strategic outcomes. And thus, when universalized, you end up in a situation where everyone may be safer relying on their interests to evaluate a reform, rather than normative concerns. Which reduces institutional change to a process fundamentally about short-term outcomes, which are guaranteed to be in conflict.

This is one reason that reformers sometimes propose blind or random change. For instance, one popular suggestion for abolishing the filibuster is to have the abolition take place 12 years down the road, when no one knows which party will be in the majority in the Senate or who will control the Presidency. The appeal of such a suggestion is obvious: it is explicitly trying to remove the short-terms strategic interests calculation from the decision-making process. By blinding the political actors from knowledge of how the reform will affect outcomes, people are bound to put more weight on their normative judgements about the value of the reform. Proposals to randomize the primary calendar offer the same benefit: with the exception of the existing winners (like Iowa and New Hampshire), strategic support based on outcome is less likely to jeopardize the success of the reform.

Of course, the converse is also true. Without short-term winners, it’s often hard to find people who will work hard for institutional change. There are definitely people out there who just want to Do The Right Thing, but they are (1) rare, (2) hard to identify, and (3) not usually backed with a ton of money or resources or hard-working volunteers. Now, I’m not trying to discredit reform or reformers here. But it’s simply true that even the normatively best reformers are usually sponsored and midwifed by the sweat of those who stood to benefit from the short-term changes to the outcomes.

2. The wide scope of institutional change. None of what was said above, of course, is either groundbreaking or limited to institutional change. It’s just as plausibly applicable to policymaking as it is to the electoral and institutional structures which provide the rules for policymaking. And as I’ve written about before, one of the fundamental blind spots of democracy is the difficulty of dealing with issues that entail long-term benefits but short-term costs.

But institutional reform is particularly tricky because any individual change has the ability to affect a whole array of policy outcomes.  And therefore, those who perceive the short-term outcomes to be against them can be quite diverse indeed. This is often carried through the reality of varying intensity of preference among political actors. Consider the filibuster: when Senators contemplate whether or not to abolish it, they are not just thinking about whether their party is currently in the majority or whether they want the current policy agenda to pass, but instead they are thinking about the individual policies most dear to them that they do not want to change.

In effect, changes to current institutional arrangements do not just have the ability to create short-term losers, they have the ability to transform everyone into a short-term loser at the same time. This partially explains why serious filibuster reform has failed to get a majority vote in the Senate the few times it has successfully been brought to the floor. Most Senators don’t want to reform the system, regardless of who the short-term winners will be, because they worry (perhaps correctly) that they will be made a short-term loser because reform will endanger the outcomes they care most about.

3. Biases against change. This is perhaps the most commonly-cited reason that institutional change is difficulty. There are at least two flavors. First, there’s the institutional structure itself: most institutional reforms are going to require supermajority support at some point or another, or at least will run into multiple veto-players along the way. A bare majority is rarely good enough for serious institutional reform. In one sense, it’s odd that we even call this a bias. Because it actually makes strong normative sense. It’s would be a strange world indeed where the a bare majority could always adjust the rules of the game willy-nilly. And that’s as true for meetings in a church basement as it is for amending a national constitution. So part of the bias is simply a different bias we have, one that favors majoritarianism.

Second, there seems to be a cognitive bias in most people against extreme negative outcomes. And therefore, people will usually take mildly worse average results to avoid uncertainty that will yield better average results but also high variance that may occasionally result in strongly negative outcomes. As discussed above, institutional reform is resides in the world of uncertainty; not only are there unintended consequences to any institutional change, but information about those consequences is very difficult to trust. Therefore, actors are likely to judge the status quo more favorably than it might deserve. Marginally improving things is great, but unless the status quo is highly untenable, individual actors often prefer to maintain it rather than choose an alternative that might results in strongly negative results.

Of course, all of this leaves the $64k question: how does institutional change actually happen? That’s best left to another post. Maybe next week.


**Please don’t mistake me for a neo-conservative here. I’m not interested in using war to build democracies. In fact, I’m not really interested in war at all. Or in trying to build democracies through destabilization of non-democratic regimes. I simply mean that we should cheer without reservation when it happens, or modestly help when asked.


2 Responses to The Difficulty of Institutional Reform

  1. […] — this is almost certainly policy preferences disguised as a normative process debate. But as I wrote yesterday, that’s not […]

  2. Recess Redux | Matt Glassman on January 6, 2012 at 3:53 pm

    […] think you can political questions are the institutional questions can be separated; as I wrote earlier in the week, institutional change is often driven (or not driven) by short-term politics. In the end, they are […]

Leave a Reply

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