In order to assign a price, it is necessary to do a hypothetical. The individual must consider the joy from the experience after the fact, act as if that is known before the fact, and then determine the willingness to pay for that experience before the fact. On the one hand, conceptually doing such a hypothetical is easy to imagine, so in that sense everything does have its price. On the other hand, economists are notoriously skeptical about the accuracy when eliciting information from people about some hypothetical choice. Often they don’t trust polling information for this reason.
President Harry Truman is known to have adlibbed, “Give me a one-handed economist…”
We have treated the procurement as a one-shot decision. In reality, decision making is ongoing, with one choice following another and the economic environment in which each choice is made interconnected by prior choices having some effect on the environment. Some of the imprecision in managing a hypothetical can be accounted for in the ongoing setting, in which case it may actually be advantageous to maintain several distinct indicators of benefit and cost rather than aggregate everything into a single indicator measured in dollars.
There are several reasons for this:
(1) Many transactions happen outside the market and those are harder to measure. In the golf example we mentioned one possibility would be to hire a caddie, a market transaction. Another alternative would be to carry your own bag, yielding no market transaction, but the work of carrying the bag gets done nonetheless. (Caddies also provide counsel to the players. If you carry your own bag, you don’t get that service.)
(2) Even if you have a process to acknowledge the non-market activity, the data for that will be softer, and there can be a type of intellectual tyranny where the hard data trumps the soft data, which can seem unimportant (when it actually matters a lot). Keeping other indicators is a way to acknowledge that some of the information is soft.
(3) There can be “externalities” that are seemingly outside the decision process but that perhaps really should be internalized. Still another alternative to a caddie or to carrying your own bag is to rent a cart. You may prefer the cart because it enables you to get around the golf course quicker. In so thinking, however, you may not consider the long term health benefit from walking. (For any one round of golf that benefit is probably nil, but over a season or an even longer time period that benefit can be considerable.) If you did adequately account for the health benefit, the decision to rent a cart or not might then well depend on how much other exercise you’ve been getting.
While this is a course on microeconomics and we normally don’t consider macroeconomic aggregates such as Gross Domestic Product (GDP) this piece from the New York Times Sunday Magazine is worth the read because it illustrates many of the same sort of measurement issues that we are considering here. Further, the piece conveys that no matter how thoughtful you might be in considering the measurement questions, to implement an actual procedure it will be necessary to make a pragmatic compromise between those constructs you’d like to account for in your measure and those things to which you can reasonably count. This means that even the designer of of the metrics who came up with the approach may nonetheless have second thoughts about those factors that are mis-measured or not measured at all. (Read Kuznets’ Nobel Prize lecture, particularly part 3, Some Implications.)
The issue of assigning a value or a cost to each component of the procurement is important, but it is by no means the only issue. The overall complexity in making the choice may be unnerving. Aggregation is a way to reduce complexity, which is good in itself, but one may be suspicious that in the process certain important things get lost. There can be a problem with right censoring or left censoring of the valuations that distorts the choice. As many students probably haven’t gone through formal procurement on their own, let’s consider an analogous situation that students are apt to be more familiar with – grading rubrics for term papers.
In a grading rubric a set of categories are selected. Performance is scored within each category on a certain scale from lowest to highest. There may also be a specification of what is expected to earn few points, more points, and the highest number of points within each category. Suppose the points within a category are on a 5-point scale, with the lowest score a 1 and the highest score a 5. There is no censoring problem at all if all the scores that are allocated are 2, 3, or 4. There can be a right censoring problem when scores of 5 are given. Suppose instead that the scale was from 1 to 6 where the performance from 1 to 5 remains as before but now a 6 goes even well beyond what was expected of a 5. When rubrics are used and right censoring is an issue, the grading scheme tends to favor papers that are broadly good over those papers that are truly exceptional in some components. Seeing that, the evaluator may then want to change the rules. But changing the scoring rules in midstream is unfair. (Left censoring is essentially the same issue but now referring to truly bad performance.)
Partly because of these censoring issues and partly simply not to be overwhelmed in light of the complexity, actually choice often reduces the dimensionality of the decision by ignoring components either where the alternatives don’t seem so different or by focusing on only those components where censoring is apt to matter. The decision maker typically produces a narrative to support the decision and this simpler version of the choice is what one is apt to hear about in the narrative.
One such area where these narratives are prominent is public policy. To illustrate, let’s consider a particular policy proposal, imposing an additional Federal tax on gasoline of $1 per gallon in order to promote fuel efficiency and the development of alternative energy sources, with part of the proceeds of the tax being rebated to consumers via a reduction in the payroll tax, and the remaining part of the tax used to subsidize research and development in alternative energy sources. Contrasting such a policy to the status quo is a form of “what if” analysis. Can you think of reasons why this proposal is a bad idea? What about other reasons under which the proposal seems a good idea?