|
Comments on Anderson's ch 4
As always, unless otherwise noted, references to Anderson are to his 2011 7th edition assigned as a text for this course.
It is often said that the legislative process is like sausage-making. Just like sausage, you don’t want to watch legislation being made. And this applies to all policy decisionmaking (or policy adoption), including presidential, judicial, and bureaucratic decisionmaking, not simply the work by legislators. Moreover, like sausage, not all policy tastes good.
Focusing on making policy, rather than choosing policy, one of the striking things is that the final policy package is often considerably different from the ideas and preferences of the original proponents (look at Senator Durbin in the bankruptcy case—Anderson: 156-62). Very often, there is a complicated transformative process between a particular policy design and the policy adopted. Moreover, recalling the point made about synergy earlier, how the policy adoption process plays out is challenging to describe after the fact and immensely difficult to forecast at the start.
The values or preferences people bring to policymaking matter. As you look at the types of values Anderson discusses (132-34), understand that these values are not alternatives (shall I be guided by my personal values or by my organizational values?). Instead, each person brings their own mix of a number of different types of values to the policymaking process.
Modes of decisionmaking
Comprehensive rationality (or, as Anderson [127-29] calls it, the rational-comprehensive model) is the textbook approach to making choices, but it simply doesn’t accurately describe human behavior. In the context of policymaking, it immediately falls apart because there is not a single decider. Even President Bush, who claimed to be the decider-in-chief, implied that others had a part in policymaking. In any event, the cognitive demands on a single decider are too great to follow the approach of comprehensive rationality—being able to understand all dimensions of a problem, identifying and weighing multiple goals, identifying all alternative courses of action and identifying the probable consequences of all alternatives, and then choosing the alternative that maximizes the realization of multiple goals.
A much more accurate description of decisionmaking generally, and policy decisionmaking which is the focus here, is the theory of bounded rationality (which is a term I prefer to Anderson’s incrementalism: in terms of a general theory of decisionmaking, incrementalism is too tied to certain patterns of outcomes, often seen in budgeting, where new decisions are frequently just minor changes from past decisions; bounded rationality can result in policy changes which are not incremental in character). Bounded rationality is anchored in a view of human beings as having many aspirations but being cognitively limited and limited in decisionmaking resources. The leading conceptual and empirical work on bounded rationality was done by an intellectual hero of mine, Herbert A. Simon, a political science PhD who spent much of his career with joint appointments in Psychology and Computer Science at Carnegie-Mellon University and won the Nobel Prize in Economics.
As you read through the following, think of a few scenarios
- Any vehicle to which you now have access is taken away, and you need to arrange transportation to school, work, shopping, recreation
- You are hungry
- You’ve been diagnosed with some disease
- Admittedly an unlikely scenario, I give you $100,000
To start elaborating the concept of bounded rationality, consider the case of a single decisionmaker who
- Has multiple, often conflicting goals: security, a good economy, environmental quality, a fair distribution of wealth (note Anderson’s listing of values [132-34)]
- Perceives a simplified representation of reality in a world that is too rich, too complicated to fully understand in all of its aspects and details (recall the observations about social constructions of reality/policy)
- Within this simplified representation of reality, becomes aware of a “problem” (which is itself necessarily a simplified representation)
- Typically, and often rather quickly, imagines an appropriate “solution” to the problem; this is vital
- Faced with any problem, we imagine what would be an appropriate solution, without which we cannot go on to determine costs and benefits of various alternatives and the probabilities of various consequences
- “Appropriate solutions” are themselves constructed by the decisionmaker (they don’t lie out there in nature)
- Some “real solutions” may exist that we never conceive because they don’t immediately appear to us as appropriate (which is, by the way, one of the advantages of coping with the complications of multiple participants in decisionmaking—two heads are better than one, and three can be really great)
- Up to a point--and this is a function of the available time, the costs of gathering information, the number of competing problems demanding solution, motivation, and real limits on human calculating abilities-- considers a limited number of seemingly appropriate alternative solutions, identifies only some of the costs and benefits of these alternatives (roughly estimating thelikelihood of realizing these costs and benefits)
- Chooses the alternative that, with its likely costs and benefits, seems reasonably good; this is Simon’s idea of satisficing (satisfactory/sufficient), rather than optimizing, in resolving the problem (as socially constructed) in light of multiple, perhaps conflicting goals
- This is a good opportunity to note that when a decisionmaker (or collegial body of decisionmakers) is faced with repetitive problems she typically fall backs on decision rules (Anderson: 140-41) and standard operating procedures, using past decisions that indeed satisficed (proved satisfying and sufficient) to apply to new cases.
Now, add in multiple decisionmakers, as is almost always the case in policymaking, where a decision can only be made when two or more participants come to agree on a policy and where each participant
- Has her own multiple, often conflicting goals
- Has her own simplified representations of reality
- Has her own idea of an appropriate solution
- Has her own imagined set of alternative costs and benefits and her own expectations of the probabilities of realizing these costs and benefits
- Has her own ideas of what alternatives satisfice
Mixed scanning (130-31) is nothing more than a normative recommendation that encourages taking more time, investing more analytical resources, and expending more cognitive effort on those problems that seem to be unusually important. Good examples here would be the Iraq surge decision by the Bush Administration and the Obama administration’s Afghanistan surge decision. There is no expectation that these problems can be solved through comprehensive rationality. They will still be resolved in a bounded rationality way, but policymakers are advised to work more diligently than they normally can in understanding the problem, be more careful not to foreclose possible alternatives, be more systematic in working to identify costs and benefits, shore up their estimates of the probabilities of various occurrences, and so forth. The trick in all this is that there are only a limited number of problems to which a boundedly-rational individual or group can give more-than-usual decisionmaking efforts.
Ways of resolving disputes: Given multiple participants in a policymaking process, how are different preferences to be resolved? Elsewhere, I have distinguished two broad alternatives: violence and politics. Politics includes Anderson’s (145-48) bargaining and persuasion, but it also includes voting, as in Congress and the Supreme Court, even though, curiously, Anderson doesn’t explicitly mention it. Command (Anderson 148-49) might be a third alternative to violence and politics; on the other hand, people obey commands either because the commander can punish disobedience by violence or because the ordered one is persuaded by the commander’s legitimate hierarchical authority or because the commander has rewards and penalties to be applied to the subordinate (which makes it a form of bargaining). Just as with violence, a total reliance on command is often dangerous and counterproductive.
Anderson (149) is quite right that policymaking typically involves multiple ways of resolving disputes. In any one policymaking case, one can often see, at different points, incidents involving persuasion, others involving bargaining, yet others including voting, and even exercises of command.
Vagueness
Anderson doesn’t explicitly talk about vagueness, and it is not apparent in his two cases in Ch 4, but very often policies as adopted involve considerable vagueness.
A favorite example (from Brown, 1979): in the Occupational Safety and Health Act, Congress directed the Occupational Safety and Health Administration (OSHA) to develop regulations that would protect workers from workplace hazards “to the maximum extent feasible.” Polyvinyl chloride is a carcinogen and it is technically possible to reduce the risk to nearly nothing, but only at a great economic cost that might drive manufacturers out of business and throw employees out of work. The issue was left to OSHA to determine, what does “maximum extent feasible” mean, technical feasibility or economic feasibility? In this case, vagueness meant OSHA had to choose between deciding whether it would focus on life and death or on jobs and production.
Policymakers, especially but not only legislators, tend to adopt vague policies for three reasons. First, they don’t have the time or interest to set out all the details. Second, they often lack the information necessary to write clearer, more specific, more directive policy. Finally, there is often a political advantage in being vague, because vagueness doesn’t signal winners and losers as clearly as does specificity; vagueness thus reduces the likelihood of threatening those with opposing views. (Note that the first two appeals of vagueness, and indeed the third as well, conform to the “bounded rationality” (aka “incremental” interpretation of problem-solving).
Vagueness in public policy is often not a satisfactory solution because it ultimately hands off policymaking authority to those who implement policy decisions or interpret them. The OSHA example, just given, is a fine illustration. The cost to Congress of leaving the policy as “to the maximum extent feasible” is that OSHA bureaucrats decide whether this phrase refers to economic or technical feasibility. And, while I don’t recall the judicial outcome, OSHA’s rule on feasibility was challenged in court, giving the judicial branch its opportunity to take a whack at safety regulations.
Odds and ends
When Anderson (126) talks about two elements of legitimacy—“how something is done…and what is being done”—he is relying on a public law distinction between procedural due process and substantive due process. And he is right: policies can be challenged on the ground that government doesn’t rightly have the authority to do something (say, favor women or ethnic minorities in government contracting) or that, in exercising a power to which the government truly does have a right, a government official has done so in, say, an arbitrary or capricious way (for example, in denying an unpopular group the right to distribute leaflets on campus when other groups are freely allowed to do so).
Sunk costs (Anderson: 128) normally refers to expenditures already made and funds that cannot be recovered. While it is psychologically difficult to cope with losses, economists make the good point that sunk costs ought to be irrelevant in making any decision. Those funds are gone. The question is always, will the benefits of a new investment going forward outweigh the costs of making that investment? Example: Construction on the Bellefonte nuclear plant in Alabama was begun in the 1970 but was subsequently halted after $4.4 billion had been spent. It is estimated to cost another $4-5 billion to finish the facility. By the economic analysis, the $4.4 billion already spent simply doesn’t matter for making a good decision on whether or not to resume construction. Been there, done that; that’s water over the bridge. The question is whether the benefits of finishing the plant are greater than the $4-5 billion completion costs.
Read constituents (Anderson: 136-37) broadly. Constituents are not only voters in a congressional district but are those people who matter politically for any policymaker (what might be called each policymaker’s political base). So, the president has to think about the interests of members of his party in Congress, the mass electorate who will determine his party’s success at the next election, his campaign donors, foreign leaders whose support he needs in world affairs, and more. As Anderson (137) points out, unelected bureaucrats also have constituents in the policy areas in which they act.
The role of public opinion in public policy is a course unto itself.
Anderson’s description of congressional decisionmaking (149-52; and also the two cases studies, 156-64) gives credence to the importance of procedural rules as outlined by the institutionalism perspective (Anderson 23-25). Not only Congress, but also the bureaucracy and the courts, and even the presidency are governed by organizational processes and routines that impact how policy is made and what policy is made.
The factors that Anderson (153-55) says constrain presidential policymaking--legality and acceptability, available resources (including time), previous commitments, and available information--apply just as much to legislatures, courts, and bureaucracies.
|
