November 20, 2014

Paradoxes and Biases of Cost-Benefit Analysis

COST-BENEFIT BREAK-EVEN FEELINGS“Instead of weighing the costs and benefits of two opposite control strategies to determine which to adopt, we seem to be unable to specify these pluses and minuses independently of a presupposed, pre-existing commitment to one of the two strategies we are supposed to be evaluating.” —M. Moffa, “Costs, Benefits and Applications of Psychological ‘Locus of Control’”

Undertaking and completing a cost-benefit analysis before making a commitment, especially in business, hiring, etc., is supposed to be the height of rationality; failure to do so, the height of folly.

I recently talked with and asked the owner of my favorite pizza shop in the world (located near my downtown Vancouver office) whether he did a cost-benefit, break-even analysis before buying the business. A pharmaceutical researcher and quite a smart guy, he said that indeed he had, and that it was repeated for the incoming new owner. (The analysis revealed that the break-even point was 250 slices per day; he then revealed that the business has been a gold mine.)

I suspect that many of the frequently, when not perennially unrented shops in the West End on Denman Street—a prime and therefore very pricey commercial real estate and tourist venue next to sightseeing gems English Bay and Stanley Park—are monuments to wide-eyed dreamers who never sat down to do a cost-benefit analysis for their over-scented candle, over-priced minuscule meat-pie and countless other businesses that quickly went under.

Hiring Cost-Benefit Analysis

Although recruiting is more about hiring than selling (except in the figurative sense of “selling” clients and candidates on the idea of doing a deal), a similar cost-benefit or break-even analysis (itself a species of cost-benefit analysis) seems warranted—at least for every hire, if not for the business as a whole, e.g., an independent agency.

Of course, the analysis will only be as reliable as the anticipated cost and benefit estimates that it depends on. Admittedly, these can be very difficult to come up with, especially when the costs and benefits include non-dollar-denominated ones, e.g., the costs and benefits of hiring a brilliant IT consultant who favors sandals without socks or of hiring the boss’s nephew (which can be very costly to you, personally, if you hire someone else).

But, there is a problem with cost-benefit analysis—a problem that is also, in some of its forms, a paradox: In many, even if not in all instances, determining the costs and benefits of strategy or decision X vs. Y requires or is contaminated by, however subtly, a prior commitment to one of them.

For example, consider the costs and benefits of thinking and deciding things logically (as opposed to emotionally or intuitively), i.e., of deciding whether to let the head rule the heart or, instead, vice versa.

A Logical Rejection of Logic?

Logically evaluating the costs and benefits of always deciding things logically requires a prior commitment to logical thinking. How likely is it that the process and form of that cost-benefit analysis will be self-invalidating, i.e., that through logical analysis it is shown that the costs of thinking and deciding logically exceed the benefits?

That amounts to paradoxically rejecting the logical grounds for the conclusion that the logical grounds should be rejected. If logic compels the conclusion that the heart should always (and not just sometimes) rule, how can that conclusion be accepted, when it is rejecting the logic that proves it?

Ditto for the opposite: a logical cost-benefit analysis that justifies the decision to always accept logical analysis (rather than even occasional intuitive or emotional decision making). Another paradox, except that this time it involves self-validating logical cost-benefit analysis, rather than a self-invalidating analysis.

The point and lesson here is that the overall utility of a cost-benefit analysis and the estimates of the associated specific costs and benefits are limited to and by the degree to which a prior commitment to or rejection of one of the strategies or decisions being evaluated among alternatives is a prerequisite for the characterization, calculation and weighting of those costs and benefits. (Yes, you may have to reread this several times to get it.)

A Harvard Bias

If the logic example seems too abstract or philosophical, a practical business-related illustration should suffice as an illustration of how bias, rather than logical necessity, can create a second kind of cost-benefit paradox: cases in which biases determine the cost-benefit analysis result in advance. Here’s one:

You have narrowed down the candidate field to two applicants. One is a brilliant graduate of a state university; the other, a Harvard graduate in the bottom 5 % of his graduating class (a fact you became aware of from an ill-advised candid Facebook posting of his). You’ve always been impressed with the Ivy League and especially with anyone who has graduated from Harvard. The consequence is that when you perform your unconscious, informal or intuitive cost-benefit analysis of the hiring choices, you will be assigning disproportionate weight to that Harvard credential, because you have a prior commitment to doing so.

In effect, you’ve biased your cost-benefit analysis in virtue and favor of a pre-existing commitment to something you should be evaluating, namely the weight (benefit or cost) to be assigned to a specific educational credential as a decision parameter. This bias effect is different from the effect of having an evidentially well-supported, widely accepted or imposed criterion or scale that will be applied in all candidate assessments. The difference here is that between judgment (using the warranted criteria) and prejudgment (using the bias).

Alternative Cost-Benefit Criteria and Measures

Consider a subtler and broader analogue in the pizza business case: There is a prior, commonplace commitment to denominate all costs and benefits in dollar-denominated, self-interest terms (as opposed to including the benefits denominated in terms of the impact of selling very popular carcinogenic sodium nitrate-laced, high cholesterol pepperoni pizza on public health or in terms of social capital gains through pizza-fueled networking and useful new friendships).

Other excluded approaches include a cost-benefit analysis based on the amount of leisure time, as a key consideration, any given decision allows if implemented. That is to say, the conventional approach excludes a criterion based on a cost-benefit analysis of the leisure-to-labor ratio by choice of strategy X, rather than Y. Conceivably, it might be “better” to choose a business model that offers more leisure time at the expense of bigger profits.

That dollar-focused predisposition will bias not only the criteria and measures of the cost-benefit analysis of the business, but also the eventual decision to start the business itself and what to offer, e.g., pepperoni pizza, whether that decision is yea or nay. In turn, that decision will depend on the final dollar-based cost-benefit balance sheet and on the prior decision as to whether to include other costs such as the full “M.O.R.E.S.” costs (my acronym)—maintenance, opportunity, replacement, evaluation and search costs associated with starting and operating the business.

Notice that this bias is different from the hiring bias that in advance favors one of the specific alternative choices in virtue of the weight, i.e., benefit, assigned it (as is the case with the Harvard bias). That Harvard illustration is more like the logic example, in which a prior commitment to choosing the specific alternative of logic over emotion or intuition likewise favors, indeed requires, the eventual choice of logic (or Harvard) as the “winner”.

It is one thing to have a bias or pre-commitment to inclusion of specific measures, standards or variables as the framework of a cost-benefit analysis, e.g., to denominate the costs and benefits in dollars only; but it is quite another to bias or ensure—in advance—the eventual decision outcome because of the selected framework elements (as was the case in both the logic and Harvard examples).

In other words,, it is one thing to depend on universally accepted quasi-“objective” measures of cost and benefit, such as dollars, but quite another to depend on idiosyncratic, prejudiced and therefore more “subjective” measures, such as personal preferences (for Harvard) when these are not accepted by supervisors or clients.

A similar point I made in my article “Costs, Benefits and Applications of Psychological ‘Locus of Control’” was that, in framing and completing a cost-benefit analysis of whether to rely on an “internal locus of control” (e.g., autonomous decision-making) or, instead, an “external locus of control” (e.g., depending on the advice of others), a prior psychological commitment to either one of these will bias the estimates of costs and benefits we have to use to determine, without bias, which of the two styles we should adopt!

For example, consider the biases and differences in the weight—positive or negative—assigned to “making group-based choices” by these two types of personalities undertaking a cost-benefit analysis of these two types of personalities.

Ask an Introvert to Objectively Evaluate Introversion vs. Extroversion with a Cost-Benefit Analysis?

That’s like asking an introvert to do a cost-benefit analysis of introversion vs. extroversion. His pre-existing introversion will bias the cost and benefit estimates for almost all the parameters in that analysis, favoring anything that shuns or otherwise excludes group involvement. For example, this will apply to the costs and benefits of group activities, such as sports or glee club (which will predictably be rated a net benefit by extroverts, but as a net cost by the introvert).

In effect, that will make the supposed cost-benefit “analysis” nothing more than a summary and statement of a foregone conclusion, namely, that introversion is, in advance of the “analysis”, the better choice (much as logic has to be preferable to emotion or intuition if it is used to determine which of the three is the best choice).

Such logically or psychologically biasing factors defeat the fundamental purpose of a cost-benefit analysis, which is help us decide, rather than to summarize what we have already through psychological bias or paradoxical force of logic pre-decided.

Notice that although the presence of bias favoring a specific cost-benefit analysis outcome is not paradoxical, it does vitiate or at least compromise the analysis, just as an introvert’s bias toward introversion invalidates his attempted cost-benefit analysis of introversion vs. extroversion.

Logic and the Heart

Paradox and bias also dog those who spurn cost-benefit analyses altogether. Consider those who refuse to believe that matters of the heart should or actually involve cost-benefit analysis, e.g., when “falling” in love (or is it jumping into it?), choosing to marry X instead of Y, or starting a family. How will they deal with those committed to cost-benefit analysis for not only all decisions, but also for all voluntary behaviors, who will predictably ask two awkward questions:

1. “Shouldn’t you complete a cost-benefit analysis and comparison of not making vs. making a cost-benefit analysis of your heartfelt decisions?”

2. “Why do you believe, how can you be sure that, even unconsciously, you aren’t doing precisely that?”

Read more in Assessment

Michael Moffa, writer for Recruiter.com, is a former editor and writer with China Daily News, Hong Kong edition and Editor-in-chief, Business Insight Japan Magazine, Tokyo; he has also been a columnist with one of Japan’s national newspapers, The Daily Yomiuri, and a university lecturer (critical thinking and philosophy).