Dubious Research (Investor Attitudes Edition)

I will let you in on a dirty secret. Appearances to the contrary, this blog isn’t about finance and markets. Its real purpose is to encourage critical thinking, but since I know a wee bit about the financial services industry, I tend to use that as the point of departure.

In that vein, it’s important to know whether information is valid. A surprising amount of research that is touted in the press, if you bother tracking down the underlying study, is questionable (common failings are small study populations and lack of controls).

Today’s object lesson is a survey discussed in Invesment News, “Investors value ethics, says study“:

Most investors prize a company’s ethical behavior over rich returns, a study found.

Seventy-six percent of investors said they would pull their money from a company that engages in legal but unethical behavior—even if that behavior brought in higher returns, a poll by Opinion Research Corp. for Pepperdine University’s Graziadio School of Business and Management found.

Two-thirds of the 482 investors polled also said that they knew about the ethical standards and practices of companies they invest in.

Two-third of the 482 investors polled also said that they knew about the ethical standards and practices of companies they invest in. About half said that corporate boards are doing a good job of ensuring that companies are managed ethically, while 42% said that boards were doing either fairly or poorly.

“Clearly, investors are looking at more than the balance sheet and sales projections when it comes to investments,” said Linda A. Livingstone, dean of the Graziadio School of Business and Management, in a statement.

“Corporate Board leadership that is centered on values and ethical behavior plays an important role in how investors evaluate options,” she added.

Anyone who knows anything about consumer research will tell you this poll mislabeled as a study is garbage. I am highly confident that if you asked retail brokers (excluding those who specialize in ethical investing), you’d be hard pressed to find one who could name a customer who had sold a stock over “legal but unethical behavior.”

So why did this study produce such misleading findings? Questionnaires have been found to be unreliable in assessing buyers’ propensity to act (consumers will consistently say they are considerably more willing to buy a product, and at a much higher price, than they ultimately are). Other analytic approaches, such as conjoint analysis, are much better suited to understanding how consumers value and trade off various product features (and in this context, “ethical behavior” is being traded off against other attributes, say dividends, P/E ratio, earnings momentum, etc.).

Similarly, in polls and other types of surveys, the wording of a question has a significant impact. If you asked, “What do you think of the job George Bush is doing” versus “What do you think of the job George Bush is doing as President,” the latter would give significantly higher scores, because the “as President” reminds respondents that it’s a tough job.

Generally speaking, if you ask questions framed along the lines of “Do you care about X” you tend to get inflated responses, because participants unconsciously want to please the questioner and the phrasing of the question implies that the questioner believes you should care about X.