One of the reasons I dump on Salar Jahedi at every possible opportunity is his constant claim to be utilizing his mighty powers as a trained Economist when he analyzes stuff. First, while all Economists are vastly arrogant, because we know everything, we also know it's best to hide it. Mr. Jahedi has not been using this advice.

His 'I took Econ 100A' moment today is a nod towards the concept of Utility Maximization.

As an economist, I know colleges are utility-maximizers. This means they opt to make the best choices given their circumstances. I trust that, as they are specialists in the field, they have come to learn what type of student is most beneficial to their program.

The SAT aids them in selecting these students. It provides additional information, which makes the colleges' decisions more accurate. If the SAT were banned from the admission process, the admission committees would not search for different students. Instead, they would search for the same students with less information to aid their search.

This is not really a straightforward utility maximization problem. Utility-maxing is typically done under conditions of perfect information, because it's impossible to perfectly max out Utility without it. This is a classic Imperfect Information problem, which means that different analytical tools have to be used, to try and figure out the optimal choice in suboptimal conditions.

Another problem with the Utility problem setup: it assumes one standard of utility. But, of course, there are all too many definitions of what would be of the most value to the University. Perhaps highest Utility comes from admitting the most intelligent. Or perhaps it's the most involved in the community. Or the best Leaders. The real problem is not as much lack of information, as Jahedi writes, but how to weight the information it gets. Take myself: I had subaverage grades, good SATs, good community service. The admission process decided my marginal utility was higher then the others... but on what weights? Does good SATs trump good grades? Where does Leadership come in? SAT scores may, if badly weighed, actually reduce total utility by letting in the wrong types of students.

Another problem: the hypothetical Utility-maxing of Admissions staff will not necessarily coincide with the Ute-max of the University as a whole. What if somebody is biased towards Boy Scouts? Or punishes people who misspell License? Irrational bias and other natural human factors will also move University Admissions away from perfect utility maximization.

And finally, how much skew is introduced by problems with the SAT? Mr. Jahedi assumes that the information they contain and the various tricks to correct for their flaws will outweight any errors, but he has no proof of this.

Mr. Jahedi's use of a utility-maxing scope, without first going through the various imperfect information and bias problems, leads him to declare that SATs must be more helpful because they're more information. Possible, but without first analyzing the skews induced by this imperfect information, he can't claim much.
Email This Post!