Over at Dispatches Michael Heath and I had a little set to over the issue of getting Congress to take scientific and technological advice more seriously. We are in agreement on the value of accomplishing that goal, but we disagree on the practicality of his proposal. I don’t want to revisit the whole debate here, and I’m not writing this for the purpose of staging an attack on him, but as a route to considering the importance of boundary conditions that constrain the making of public policy. For example, Roger Pielke, Jr.’s Iron Law of Climate Policy, which says that “[w]hen policies on emissions reductions collide with policies focused on economic growth, economic growth will win out every time” is a boundary condition. Any proposal that is based on the public accepting substantial short-term economic costs is a non-starter. Whether we believe that there is really no choice in the matter, that we have to accept short-term economic costs, is irrelevant–we always have the capacity to make these allegedly impossible choices, and the certain prediction is that we will. (Or, if not certain, then certainly the prediction with the highest probability of actually occurring.) The underlying issue here is the unpleasant reality of the constraints these boundary conditions impose–the way they make accomplishing a desirable goal, such as getting Congress to take scientific advice more seriously, even more difficult than one might initially expect.
Here is Michael’s proposal, slightly paraphrased, but mostly using his words:
Congressional committees should by law be required to develop legislation with input by a scientific agency like the former Office of Technology Assessment. That agency would charter a study by experts specific to the particular issue to present a proposal. The congressional committee than develops its own prescriptions, but is forced by law to highlight deviations from the expert proposal and justify those deviations.
As much as it would be desirable to have Congress take more notice of technical expertise, I believe this proposal faces at least insurmountable boundary conditions. One I noted in our discussion over there, while the other occurred to me last night while watching the Detroit Red Wings play a particularly uninspired game at Joe Louis Arena (thanks guys, that was my first time watching you live–I want my $25 back).
Boundary Condition 1: Congress Is Immune to the Law
Congress is the lawmaking authority in the United States. Roughly speaking, we can say that it is “the government” (whereas we would call the executive, “the state”). All legislative authority is vested in Congress, so Congress as an institution isn’t constrained by law. As the former Capital Hill staffer I was not-enjoying the game with stated unequivocally: “Force Congress? That cannot be done.”
An on-going example of this is the so-called “debt ceiling,” a congressionally mandated limit on how much debt the U.S. government can incur. A simple thing happens whenever we bump into it–Congress raises the debt ceiling. It is a “law” that effectively is entirely unenforceable, because ultimately Congress determines the U.S.’s level of indebtedness, and it can just revise the law whenever it becomes inconvenient not to do so. (Yes, the president can veto, but he never does.) A mandate for Congress to do something is no mandate at all. In an influential 1995 paper, political scientists Elinor Ostrom and Sue Crawford tried to create a general framework for analyzing institutions (rules, procedures, norms, mores, and the like). They distinguish five distinct attributes that can be used to categorize all institutions, one of which is the “OR ELSE” attribute;
OR ELSE is a holder for those variables which define the sanctions to be imposed for not following a rule.
That seems simplistic at first glance, but they go on to emphasize the importance between institutions that have an “OR ELSE” and those that don’t , and how it crucially affects our prediction of the likely behavior of those the rule is designed to command;
The content of the OR ELSE affects the very nature of a rule. Rowe (1989) discusses the difference between a speed-limit law with minor sanctions and a speed-limit law with a severe penalty as the sanction. The prescription is the same. For both laws, the severity of the sanction in the OR ELSE is the only difference, but expected behavior is different.
The absence of the ability to impose any sanction on Congress, to compel them to obey the law, means we cannot expect that there would be any change in their behavior if they did pass such a law, insofar as actual compliance is concerned. They might use the experts’ recommendations as a smokescreeen, a way to cover their own asses with unpopular bills they would have passed anyway (legislators are quite skillful at directing the blame toward others–it’s common for members of Congress to point the finger of blame at executive agencies for how laws are implemented, even when the implementation is carefully constrained by the laws Congress passed), but there is no reason to predict compliance.
Boundary Condition 2: The Ritualizaton of Rule Compliance”
Assume we did compel Congress, through passage of a constitutional amendment (and set aside the pesky issues of how it would be enforced). The result would be what I call a ritualization in Congress’s approach to complying with the rules. That is, compliance would become a strictly ritualistic affair, doing only what is necessary to achieve minimal legal compliance, rather than working to promote the substantive purpose of the law.
This is what has happened with Environmental Impact Statements. Originally conceived of as primarily a way of helping us understand the impact of the actions we take, particularly in regards to construction projects, so that we could make more informed decisions about whether to undertake them or not, EIS’s almost immediately became little more than a costly procedural hurdle on the way to getting approval for the project as initially envisioned. Environmentalists, who have the most reason to appreciate the concept of the Environmental Impact Statement, regularly complain about biased selection of assumptions that go into defining the costs and benefits of proposed projects. Except for when they are successfully challenged in court, the EIS is probably the least useful of all our environmental legislation.
My favorite example of ritualistic compliance was one that I engaged in some years back. After leaving the University of Oregon for my post-doc at Indiana University, I was due some bonus money for teaching at Oregon. The rules forbade just getting a check for the amount–I could spend it on research, conferences, books, or equipment. I needed a color printer and a zip drive for the computer simulations my research team was running. Sure, I could buy that, but it would belong to the University of Oregon, and I would have to return it to them when I was no longer employed, i.e., as soon as I bought, since I was in fact already no longer employed by them. Solution? Loan it to my research team leader who loaned it back to me. Or in other words, go down to the computer shop in Bloomington, Indiana, buy it, announce the double-loan process, and get reimbursed. It was a great lesson in bureaucracy–if you can find the right formal designation for your action, you can do it, even if the substance of it is strictly forbidden.
It would require a gross underestimation of Congressmembers to assume they wouldn’t figure out the ritualistic approach to meeting the requirements of highlighting its deviations from the recommendations of the experts. The people who work on Capitol Hill are very clever, very devious, bastards. I don’t say that from a generalized anti-government perspective, but from a professional perspective. Passing legislation is blood sport, and the real gladiators aren’t so much the legislators themselves (well, some of them are, but not all) but their legislative directors and legislative assistants. It’s not the classic deliberative process promoted in our civics classes, but a process of wheeling, deeling, and screwing other people over when necessary (and sometimes when it’s not). The best way to learn how Congress really works is to find a former legislative director and take him out for drinks.* If this rule was put in place, every effective legislator would have someone on staff who knew how to write the boilerplate “deviation” language. And it would never be hard for them to find alternative experts (or “experts”) to cite as justification.
The unfortunate reality is that it would be child’s play for these people. Writing laws is their specialty. They’re experts in the careful selection and construction of legal terminology, and they’d have no trouble playing this game. The outcome would be that, like an EIS, the deviation language would become ritualized and stripped of any substantive purpose. In other words, the proposal would not actually achieve its policy goal, but would turn out to be purely symbolic.
It might fairly be offered what alternative I would propose. Nothing direct. I don’t believe this is a problem that can be addressed directly. I think there are only two plausible proposals, both of which are very indirect. The first is to continue pressing for better science education in the U.S. The best way to do that is to tell the public it’s about being economically competitive with other countries. That’s an issue that has some motivating power. I support this approach in my own small way by taking science seriously in all my classes, frequently making references to evolutionary theory in a matter-of-fact tone, and requiring our majors to take methods and statistics courses. (I take pride in having instituted the rules requiring both those courses, and I take pride in teaching the methods course, where I do my best to teach them to think empirically rather than normatively). The second proposal is to move toward nonpartisan redistricting boards whose mandates are to maximize the number of competitive congressional districts. I think the gerrymandering of districts has done incalculable harm by promoting the election of non-centrist candidates in both parties. The problem is especially bad in the GOP because they tend to have highly anti-science candidates, whereas the Democrats tend to have scientific ignoramuses who have a mystical reverence for science. That’s bad enough, but nowhere near as bad as what the GOP offers. And as important, narrowing the ideological distance between legislators puts them in a position where they are more likely to seek something approaching consensus, and when they struggle with that, scientific claims provide a sort of focal point. The legislators don’t have to be knowledgeable about science, or even particularly enthusiastic about it, to converge on such a point, because science has a cachet and respectability that all except the ardently anti-scientific tend to recognize.
Is that all a slow process? Yes, indirect approaches often are. That’s why direct approaches seem so appealing–they seem quicker and more clearly targeted toward the goal. Unfortunately that perception is often incorrect. Direct approaches are more easily deflected because they are more easily seen for what they really are. The approach I suggest isn’t easily deflected because it’s not just a new rule in the game, it changes the nature of the game–particularly, it changes the nature of the players. If we don’t change the nature of the players–the anti-science types–there is no effective solution.
*At the hockey game, my friend told me the story of a nice young lady who got a job as a staffer on the House Rules Committee, the center of Capitol Hill viciousness because it determines a bill’s prospects for amendments. Allowing amendments means the bill’s effects could be changed substantially, or it could simply be defeated through the attachment of killer amendments (and opponents will look to attach killer amendments to a bill). He warned her against taking the job, and several months later she’d become a chain smoker, had developed digestive problems, and couldn’t sleep. It’s truly not for the faint of heart, or those of a kind and gentle disposition.