Drunk people are more impatient and less generous

Understanding the effects of alcohol on our behaviour is important for policy. Do people make more risky decisions when under the influence? Are they more impulsive? Are they more, or less, affected by systematic biases? Answering these questions would help with designing appropriate harm-minimising policies, or at least better assessing the costs and benefits of such policies.

But assessing the impacts of alcohol is also hard, and not helped by garbage research like this. In contrast, a paper last year (pdf) by Luca Corazzini (University of Padua), Antonio Filippin (University of Milan), and Paolo Vanin (University of Bologna) takes a much more robust approach involving lab experiments, different from the traditional observational or field experimental approaches.

Why is it important to find a robust approach? The authors explain:

First, empirical studies of alcohol intoxication based on field data, whether collected from directly observed or from self-reported behavior, typically suffer from self-selection into drinking... Any correlation between blood alcohol concentration and certain behavioral traits may reflect a true causal effect, but could also stem from different propensity to drink alcohol by individuals with those traits.

Second, and relatedly, individuals usually choose at the same time whether, when, where, with whom and how much to drink alcoholic beverages. This means that it is usually hard to disentangle the effects of alcohol from those of the context in which drinking takes place.

Third, the behavioral effects of alcohol intoxication are partly pharmacological and partly triggered by a psychological reaction to the subjective perception of being under the influence of alcohol. Disentangling the two effects requires independent variations of actual and perceived blood alcohol concentration (with implied relevant misperceptions).

In other words, people who drink may be systematically different from those who don't drink, the drinking context matters, and people may behave differently not because of the alcohol itself, but because of how they think they should be behaving under the influence of alcohol.

The paper uses a neat experimental design to get around these problems. They had three experimental groups: (1) received no alcohol and was never told the experiment had anything to do with alcohol; (2) drank no alcohol before the experiment and didn't know whether they had been given alcohol; (3) drank alcohol before the experiment  but didn't know whether they had been given alcohol. Comparing the first and second group gives an indication of the placebo effect of alcohol on behaviour, while comparing the second and third groups gives an indication of the pure pharmacological effect of alcohol on behaviour. They then ran their subjects through a battery of different lab experiments to identify their risk tolerance, impatience, and pro-social behaviour.

What they found was interesting:

Concerning risk preferences, after controlling for optimism, the willingness to pay and other subjective controls, we only detect a marginal positive effect of alcohol intoxication on risk aversion for female subjects.

On the contrary, we find a strong pharmacological effect of alcohol consumption on time preferences: it makes subjects more impatient. The pure impact of alcohol consumption on time preferences remain substantially large even after taking into account its interplay with subjects’ risk attitude. In this respect, net of the pharmacological effect of alcohol intoxication and in line with previous studies, we detect a negative and significant relationship between impatience and risk aversion.

Finally, concerning altruism, our results suggest that alcohol makes subjects more selfish, as we observe a negative and significant relationship between alcohol intoxication and donations to NGOs.

Now, their results might be sensitive to some selection bias, in that the participants in the second and third groups knew that the experiment had something to do with alcohol (which was in the advertisement for participants), whereas those in the first group did not. The no-alcohol group was slightly older and had a much higher proportion of women. Which makes the results on risk aversion a little shaky. The sample size was a little small too - an opportunity for replication beckons (though I suspect your institutional review board would take some convincing).

What do the results tell us (other than that drunk people are more impatient and less generous)? The authors suggest that "alcohol intoxication makes decisions more determined by emotions and less by deliberation", which should not be a surprise to any of us, and provides good reason for policy to place some moderate restrictions on alcohol availability.

[HT: Steve Tucker]

Why study economics? Even if you won't be an economist edition...

Last year there was an interesting article in the Journal of Economics Education by Thomas Carroll, Djeto Assane, and Jared Busker (all University of Nevada, Las Vegas) entitled "Why it pays to major in economics" (ungated earlier version here). In it, the authors use data from the 2009-2012 American Community Survey to look at the earnings of degree holders, by major. Importantly, because of the size of the dataset (millions of observations) they are able to control for not only the demographic characteristics of each person, by their job type (a combination of occupation and industry) and location. That means that their results show the relationship between college major and earnings for people in the same job.

They find that overall both male and female economics majors earn more than other majors (which other studies have found too), but importantly:

When we added occupational controls, we found that the average male Bachelor of Arts in economics major earns 8.64 percent more than his non-economics-major counterpart working in the same type of job. The advantage for female economics majors is 5.37 percent more than their counterparts of the same age, ethnicity, and job, who have different majors...

Essentially, about two-thirds of the bachelor's degree premium for economics major can be attributed to the type of job economics majors perform, and about one-third is a premium that economics majors earn over other workers within the same job.

In other words, economics majors tend to be employed in jobs that pay more on average than other jobs, but this only explains two-third of the extra earnings that economics majors receive. So, even if you aren't employed as an economist, these results suggest that there is an earnings benefit to having an economics undergraduate major (the authors also show there are benefits for economics majors who have done further study as well).

Is there something intrinsic about economics majors that lead to these higher earnings? Carroll et al. quote this earlier paper by Black, Sanders and Taylor (ungated here):

In a good undergraduate economics program, students develop an ability to think critically: They gain broadly applicable analytic and quantitative skills that improve decision making in a wide range of tasks. In short, it may be that economics majors are better trained than many other majors in skills that have returns in the marketplace.

I would agree with that - economics does provide students with skills that are both widely applicable, and in demand. Of course, that attributes causality to the relationship between economics major and earnings, which may be problematic. Maybe it is that economics students are naturally more intelligent, harder working, or just generally better employees and are rewarded with higher wages as a result? More research needed, but I would still argue that the case for completing an economics major is very strong.

Read more:

• Why study economics? PhD edition...
• Why study economics? It could make you rich...
• Why study economics? Sheepskin effects edition
• Why study economics? NZ graduate earnings edition
• Why study economics?

Do we need a price-comparison-website comparison website?

Last month The Economist reported on price-comparison websites (where consumers can compare prices between different providers of insurance, for example, or electricity - like Powerswitch in New Zealand, or other products and services). The article made for interesting reading:

Comparison sites, whether for insurance or something else, introduce a new layer of costs, including their own splashy advertising campaigns. In theory, competition in the market for comparison sites ought to keep those costs down. But in a recent paper, David Ronayne of Warwick University argues that consumers often lose out from comparison sites. They earn a commission for each shopper who uses them to buy insurance. That referral cost is incorporated into the price the consumer ends up paying. If the increased costs outweigh the saving the comparison enables, consumers end up worse off.

For instance, suppose some consumers are loyal to a single comparison site, and do not use any others to compare prices. The lucky website can crank up its referral fees, safe in the knowledge that insurers must pay up if they want access to its captive market. Those fee hikes are then passed on to consumers in the form of higher premiums.

Having a few loyal consumers would not be enough for a price-comparison website to have a high degree of market power. It would need a lot of loyal consumers - enough so that other price-comparison firms would find it difficult to make a profit, and so would choose not to enter the price-comparison website market. In other words, having a high number of loyal consumers constitutes a barrier to entry for other price-comparison website firms. It is hard to say what proportion of loyal consumers you would need to effectively lock out potential competitors, but it may be a lot. Are consumers likely to be loyal to a single price-comparison website? It's hard to say - by definition by using a price-comparison website they aren't loyal to their insurer, so why would we believe they would be loyal to a single price-comparison website?

Anyway, let's assume that there are a lot of loyal consumers, which provide a single dominant price-comparison website with market power. The higher the proportion of consumers who use that site for comparing prices, the more market power the site will have. Firms with market power are able to raise prices above marginal costs - in this case the price of referral fees to the insurance companies (or electricity retailers, or whoever). This leads to higher costs for those firms, who respond by raising prices.

At the extreme, if there is only one price-comparison website, they will (almost) have a monopoly on providing price information to consumers (I say almost because consumers could shop around themselves, albeit incurring higher search costs in the process). The single price-comparison website could increase the referral fee greatly, making consumers much worse off in the process.

How could this problem be tackled? Government could intervene in the market for price-comparison websites through price regulation (regulating the referral fees that price-comparison websites can levy on firms), or the government could run the price-comparison website itself (then it can set a low referral fee, or even none at all if the taxpayer is happy to subsidise the effort). Or the government could try to increase competition between different price-comparison websites, although that will be tricky if consumers are very loyal.

An alternative solution is to turn the price-comparison websites own business against them, and create a price-comparison-website comparison website. Then consumers could investigate to determine which price-comparison website will give them the best deal. That should increase competition between the price-comparison websites, lowering their referral fees and the prices for consumers. But would the price-comparison website comparison website then charge referral fees to the price-comparison websites? Maybe that just shifts the market power problem one step further up the chain.

Your workload impacts your grade more than you think it does

One of the most well-established cognitive biases is that people are over-confident in their own abilities. We believe that we are better than we are. This holds for students' expectations about grades as well - on average, at the beginning of a course students expect a higher grade than they end up receiving.

A recent paper by Belayet Hossain and Panagiotis Tsigaris (Thompson Rivers University) takes a closer look at students' grade expectations (sorry, I don't see an ungated version). The authors collected data on grade expectations from students across six semesters of a second-year statistics for business and economics course. Their overall test is whether students expectations are rational (unbiased, and incorporating all available information relevant to grades), which is a pretty high standard to meet. Not surprisingly, they find that expectations are not rational. However, it is some of the other results that are most interesting:

Like other studies in the field, we found that students are mostly overconfident. The unbiasedness hypothesis is rejected for most students. Expectations change sluggishly during the semester. The study suggests that students' expectations improve as more information regarding actual performance becomes available. Hence, expectations carry valuable information, and adjustments are made... Course load has a negative impact on a student's final grade after controlling for grade expectations.

That last point is important. It suggests that students who have a heavier course load are more over-confident about their grades than students with a lighter course load. Which has been my experience too - even top students over-estimate their ability to maintain top grades when over-loading their studies. All of which suggests that you should be more realistic about the degree to which study workload is going to have a negative effect on your grades.

Finally, a couple of gripes about the analysis in the paper. First, there are six semesters of students, so observations of students within each semester are not independent of each other. So, when pooling the sample the standard errors should probably be clustered at the level of the semester (Jeremy Miles gives a good, though technical, explanation of why here). Not clustering the standard errors leads to results that are more likely to show up as statistically significant than they 'should'. Second, there are likely to be unobserved differences between semesters, so semester fixed effects should probably have been included too. Omitting these fixed effects may lead to biased coefficient estimates in the models. The combination of these two problems should lead us to doubt the robustness of the results, even though they are mostly similar to those in other studies.