Ordinary Squares

Although Beyonce thinks you should put a ring on it, new research in the Journal of Marriage and Family suggests that a key ring might just be enough.

Besides love and preferential tax treatment, some of the strongest arguments for marriage suggest that married people are happier, healthier and less depressed.  But an analysis by Kelly Musick and Larry Bumpass finds that marriage might not be the mechanism at all.

In their new article, Musick and Bumpass argue that most studies of the subjective benefits of marriage compare married individuals to single ones.  Doing so, however, conflates the effects of marriage with the effects of forming a partnership.  Instead, the authors analyze data from more than 2,000 individuals to compare the benefits of cohabitation with the benefits of marriage.

The research looks at self-reported measures of happiness, depression, health, self-esteem and social ties (both with family and friends) between three groups: single individuals who remain single, single individuals who transition to cohabitation, and single individuals who transition to marriage.  By comparing single individuals who transition into marriage with single individuals who transition into cohabitation, we can better disentangle the benefits of marriage from those of simply forming a partnership.  There are several reasons to think that the transition to marriage or cohabitation could lead to a change in these subjective measures, ranging from the social support of partnerships (both for cohabitation and marriage) to the institutionalization of the relationship (available only through marriage).

The study suggests that many of the benefits we often ascribe to marriage (e.g., increased happiness, better health and stronger social networks) likely result from partnership formation, rather than marriage.  On the whole, individuals who transitioned to marriage fared no better than individuals who transitioned to cohabitation.  Where the authors do uncover differences between marriage and cohabitation, they write, “marriage was not always more advantageous than cohabitation: The married fared better in health than cohabitators, but the opposite was true of happiness and self-esteem.”  They also find that changes in health, happiness and wellbeing are most pronounced for those who formed unions recently – in the last three years – than for those who formed unions more than four years ago.

Although they equivocate on the implications of the research, as all good social scientists do, their results seem to channel a modified version of Beyonce: If you like it then you should’ve put a key ring on it. 

(Get it?  Because both married and cohabitating people share the same keys to their apartment … Does that even make sense?  Is that funny?)

Beware, active Facebook users!  The cost of all those friends just might be your happiness, according to researchers at Utah Valley University.

A new study published in the Journal of Cyberpsychology, Behavior and Social Networking (yup, that’s right!) reports that active Facebook users are more likely to report themselves unhappy and unsatisfied relative to their friends. 

The research draws on a survey that asked undergraduates when they joined Facebook and how many hours they spend on it each week.  It then asked how much they agreed with the following three statements:

·       Many of my friends have a better life than me.

·       Many of my friends are happier than me.

·       I believe strongly that life is fair.

And you wanna know what makes people feel unhappy and unsatisfied relative to their friends? Facebook.

Frequent Facebook users are more likely to report that their friends have better lives than they do and that their friends are happier than they are.  According to the research, as the number of Facebook friends on your friend list who you don’t personally know increases, so too does the likelihood that you report that others have a better life.  Likewise, early adopters – those who have been using Facebook for the longest period of time – are more likely to feel that others are happier than they are and that life is unfair.

So, what should we make of these findings? The authors suggest that frequent users recall recent Facebook content that lures them into thinking others are happier. The researchers write, “looking at happy pictures of others on Facebook gives people an impression that others are “always” happy and having good lives, as evident from these pictures of happy moments.”  The more frequently you log onto Facebook, the more frequently you are reminded of all the fun activities your friends are doing.

But I’m skeptical of this interpretation.

Instead, I suspect that the causal arrow runs the other way.  It’s not that using Facebook frequently makes you unhappy; it’s that unhappiness makes you use Facebook more frequently.  Running the causal arrow the first way, as the study authors do, suggests that scrolling through pictures of my friends’ weddings, children and island vacations causes me to perceive myself as unhappy.  Running it the other way, as I suspect is more appropriate, suggests that the only reason I’m scrolling through your pictures is because I’m already unhappy.

To be fair, though, maybe the causal arrow doesn’t really matter.  I mean, I don’t need to log into Facebook to see your wedding pictures and island vacations to make me unhappy.  I actually get sad simply thinking about the number of hours I’ve wasted on Facebook …

Victories on the college football field could lead to losses in the classroom, according to a new study by researchers at the University of Oregon.  Tracking the relationship between the winning percentage of the Oregon Ducks football team and the academic performance of non-athlete students, Jason Lindo, Glen Waddell and Isaac Swenson find that the better the University Oregon Ducks perform on the football field, the worse Oregon students perform in the classroom.

This new working paper examines the academic performance of non-athlete students at the University of Oregon over an eight-year period.  It focuses on the GPA of students during the fall term – the height of football season – to investigate whether the winning percentage of the school’s football team is associated with changes in the GPA of Oregon students.

Between 1999 and 2007, the success of the University of Oregon football team varied substantially.  In their best season, the Ducks won 92 percent of their games; in their worst season, the Ducks were victorious only 45 percent of the time.  Drawing on this year-by-year variation in the success of the football team, the authors find that the more the football team won, the farther the GPA of non-student athletes fell.

The authors draw an interesting comparison to well-known research on the effect of SAT scores on college GPAs.  Among males, a 100-point increase in a student’s SAT score is, on average, associated with a 0.16 in his GPA in college.  By comparison, they report that a 25 percent increase in the winning percentage of the Oregon Ducks – an additional three victories in a twelve-game season – results in a decline in the average GPA of about 0.04 points - equivalent to a 27-point decline in SAT scores.

The effect seems substantively small, at least on first pass.  However, the authors report that the effect is concentrated among male non-athlete students, resulting not only in a decline in academic performance, but a rise in the gender gap at the University of Oregon.  On average, women at the University of Oregon maintain a GPA 0.18 points higher than men.  The focused effect of football victories on the academic performance of male non-athlete students widens this gap.

And it’s not just academic performance that suffers.  In a follow-up survey with students at the University of Oregon, the authors report that success on the football field leads male students to disproportionately drink more, party harder and study less.  Although I’m skeptical of survey findings posed as abstractions (e.g., “Compared to a loss, when the football team wins I tend to …”), it’s hardly surprising that students report more celebrating and less studying after football victories.

So, if you were scouring for another argument to back your cocktail-party rant about the negative impact of big-time athletics on the missions and success of higher education in the United States … You’re welcome. 

If, like most readers of Ordinary Squares, you picked up last Sunday’s New York Times, you might have skimmed through an OpEd investigating whether economists (and economics students) engage in less pro-social behavior.

The OpEd cites recent research from economists at the University of Washington.  Using administrative data from the registrar, Yoram Bauman and Elaina Rose test whether enrolling in college-level economics courses changes the likelihood that students voluntarily donate to public-interest funds. 

During their online registration process, students at the University of Washington have the option of donating $3 to two different organizations – the Washington Public Interest Research Group (WashPIRG) and Affordable Tuition Now (ATN). Over the study period, about 25 percent of UW students donated to WashPIRG and 34 percent donated to ATN. But does the completion of an economics course or two make students less likely to contribute to these public interest groups?

The research finds that, on the whole, econ majors are less likely to contribute to WashPIRG and ATN than non-econ majors.  Even before they’ve stepped foot in the classroom, students that chose economics are less likely to give than students that choose other majors.  And there’s no effect of coursework on the econ majors.  After taking both an Introductory Economics course and an Intermediate Economics course, the likelihood that an econ major will donate to WashPIRG or ATN holds steady at about 5 percent.  This finding suggests a selection effect – that students who are less likely to give are also more likely to choose economics as their major. 

For non-majors, though, the effect of economics coursework is different.  After taking an Introductory Economics course, non-econ majors are significantly less likely to donate to both WashPIRG and ATN.  (The results are somewhat more ambiguous for Intermediate level economics courses.) 

What underlies this relationship?  What about Introductory Economics makes non-econ students less likely to engage in pro-social behavior, as defined by donating to WashPIRG and ATN?  The authors can’t be sure.  “Further research is needed to determine the cause of this effect, e.g., whether it results from exposure to certain concepts – such as homo economicus, the prisoners’ dilemma, or the invisible hand – or from exposure to certain people – namely, the faculty and students in the economics department – or from some other cause.” 

I bet taking Sociology courses makes them more likely to donate.

Political scientists often focus on the role of economic indicators in predicting election outcomes.  When the economy is doing well, the chances of incumbents winning reelection rises; when the economy is faltering, challengers often fare better.

But how do voters evaluate the economic fortunes of the country?  There are a myriad of economic indicators (e.g., the unemployment rate, income growth, GDP, etc.) measured at different geographies (e.g., nationally, statewide, etc.) for different categories of citizens (e.g., the wealthy, the poor).  In recent years, income inequality has emerged as one of the favorite economic indicators.  What role has rising income inequality played assessments of the national economy?

Not much, according to an exciting new article by Daniel Hopkins, a Georgetown political scientist.

Hopkins begins by looking at whether the rich and the poor differ in their assessments of the national economy. We might expect that wealthy Americans, who have experienced substantial income growth, are more likely to positively assess the national economy than poor Americans, who have experienced stagnant income growth.  Especially as the income of the poor has fallen relative to that of the wealthy, we might expect low-income Americans to assess the economy in an increasingly negative light.

Hopkins examines more than thirty years of individual-level data from the Michigan Survey of Consumer Attitudes.  As it turns out, rich and poor Americans report very similar perceptions about the state of the economy.  “Whether wealthy or poor,” Hopkins writes, “Americans are reporting the same perceptions when it comes to overall economic performance.  Given rising political polarization by income … and given the strong likelihood of partisan differences in economic perceptions, this near-unanimity is all the more striking.  When asked about the nation’s economy, Americans at various income levels give answers that are quite similar.” (13)  Even as income inequality has grown, the rich and the poor don’t diverge much in their assessments of the nation’s economic health.  The finding suggests that poor Americans are not using their declining fortunes relative to the richest Americans to evaluate the state of the economy.    

If Americans are not looking to income inequality in assessing the economy, what economic indicators do they use?  Recent research by Larry Bartels suggests that low-income voters are more responsive to gains among high-income Americans.  When the incomes of high-income Americans grow, low-income voters positively reward politicians, likely misperceiving economic growth at the top of the income distribution for economic growth more generally.  But Hopkins reports something different.  He finds that the strongest predictor of economic assessments is not the income growth at the top, but rather the income growth at the bottom.  The income growth of the bottom 20^th percentile is most strongly tied to perceptions of the national economic well-being.  Most interestingly, he finds that this holds for low-, middle- and high-income Americans.

This paper throws a curveball into the growing consensus about the importance of income inequality for our economic, social and political life.  It suggests that rising inequality does not influence how the rich and the poor view the state of the national economy, a variable that is considerably influential in shifting electoral outcomes.  The rich and the poor report similar assessments of that nation’s economic health, and rising inequalities have not made the poor more pessimistic about the state of the economy.

I reached a milestone this year: I’m halfway to the age where I can collect social security.  To be honest, I don’t know much about how social security works – and according to a recent experiment by a couple of Ivy League economists, neither do many Americans.

In a working paper from the National Bureau of Economic Research, two economists conducted an experiment to see whether providing information to elderly workers about social security changed their labor market patterns.  Because the benefits derived from social security rise with the age at which the benefits are claimed, elderly workers are incentivized to continue their workforce participation.

 But people don’t always understand the incentive structures surrounding these types of decisions.  On one hand, the rules governing social security payouts may be straightforward, but they may also require a substantial investment of time.  (This reminds me of my recent efforts to pick a 401(k) plan.  It’s not that the plans were complicated to understand; they were just boring to read through.) On the other hand, it may be that the rules are actually pretty complex and indecipherable to the lay worker.  Existing priors may also bias workers’ decisions, causing them to misperceive the incentive structures.

 The economists conducted a simple experiment.  They mailed one group of workers a short brochure highlighting key provisions of Social Security.  The brochure explained how benefits rise with the age at which the benefits are claimed, and gave several examples of hypothetical workers.  The brochure invited recipients to participate in a fifteen-minute on-line training session, as well.  The control group, on the other hand, was not sent any information.  If misperceptions about the benefits of social security change our employment choices, then the treatment group should be more likely to extend their participation in the labor market to maximize their benefits.

 Sure enough, participants in the treatment group were more likely to continue working.  One year after the experiment, 74.4 percent of the control group was still working.  In the treatment group, on the other hand, 78.6 percent of respondents reported working.  These differences are significant, indicating an effective treatment.  Information helped individuals understand the incentive structure behind social security.  The authors call this phenomenon “knowledge updating,” noting that basic, legible information resulted in behavioral changes for elderly workers.   

 More interestingly, though, is the finding that the effects of “knowledge updating” were concentrated among female workers.  Among women, the number that reported working one year after the experiment differed by 7.2 percentage points between the treatment and the control groups.  Despite the intervention, there was no statistically significant difference for men.  Although their hypothesis is, at best, speculative, the researchers suggest that elderly female workers were more likely to misperceive the benefits of continued employment.  Born in an earlier era, they were likely to be secondary earners in their households, and as a result, they may have viewed their continued employment as irrelevant to the final households benefits.

The authors conclude that, “a relatively mild informational intervention can have important impacts on the labor force participation of older individuals.” Their research underscores the importance of presenting clear, concise information to individuals navigating opaque incentive structures.  

After the election of Barack Obama, many commentators declared the end of racial politics in the United States.  They heralded the election of the first African-American president as proof that voters no longer considered race an important factor in selecting their preferred candidate.

Now forget the ecological fallacy of that argument and consider a handful of new studies suggesting that his race may have cost Obama a significant share of the vote.  One innovative paper from Harvard graduate student Seth Stephens-Davidowitz uses Google Insight to identify areas of the United States where racial animus was high.  He suggests that Obama significantly underperformed in these areas when compared to the vote share received by Kerry four years earlier.

Another interesting analysis of the effect of racial attitudes on vote choice in the last presidential election comes from a forthcoming article in Political Psychology.  There, Brian Schaffner combines measures of racial salience and racial conservativism to ask about the importance of a candidate’s race to voter preferences.  To measure racial salience, the author uses a survey question asking respondents to rank the importance of six pieces of information about a candidate to their vote choice.  The question asked, “If you had to vote in an election but did not know any of the candidates competing, which pieces of information would be most useful for helping you decide who to vote for?”  The six pieces of information were: the candidate’s party; candidate’s occupation; which candidate your friends support; whether the candidate was endorsed by the local newspaper; the candidate’s race; and the candidate’s gender.  When a respondent listed “candidate’s race” high on their list, Schaffner takes that as a measure of racial salience. 

The author then looks at racial conservatism using a question on affirmative action.  Respondents were asked to report their level of support affirmative action policies in the United States - strongly support, somewhat support, somewhat oppose or strongly oppose.  Putting these together, Schaffner looks at the joint influence of racial salience and racial conservatism on the likelihood respondents reported voting for President Obama in the 2008 election.  

On its own, racial salience is not particularly important.  Among white voters, individuals who reported race matters were slightly less likely to vote for Obama.  However, the effect of racial salience varies substantially according to the position respondents took on affirmative action.  Among supporters of affirmative action, the race of a political candidate was not an important factor in voting for Obama.  Supporters were equally likely to report voting for Obama if they cared about race or if they didn’t.

However, racial salience mattered quite a bit for opponents of affirmative action.  In his models, Schaffner reports a very low predicted probability (0.06) that an opponent of affirmative action who held race to be important voted for Obama.  Among opponents of affirmative action who reported that race was not important, Schaffner reports a much high predicted probability (0.36) that they voted for Obama.

These findings – that the racial identity of a candidate impacts the vote choice of affirmative action opponents, but not affirmative action supporters – may not surprise you.  But on account of these findings, Schaffner estimates that Obama lost about three percent of the vote nationwide. (Notably, this estimate is similar to the estimate other social scientists, including Stephens-Davidowitz, report as well.) He won handily in 2008, but he might face a tougher fight this time around.  That three percentage point margin could very well make the difference in a close election.

Everywhere you look on TV, it’s sex, sex, sex.  Rachel and Finn are almost ready to have sex in TV’s most painful courtship.  Kurt’s fighting off Sebastian in his quest to bed Blaine.  And Zooey Deschanel, star of the maddeningly unwatchable “comedy” New Girl, seems barely able to talk about anything else.

While these plotlines make for dull entertainment, they may be good for our sexual health, according to a trio of Ohio State researchers.  In a study released this summer, Emily Moyer-Guse, Adrienne H. Chung and Parul Jain suggest that watching conversations about safe sex on television helps individuals model those behaviors in real life.  Watching television characters – and especially characters with whom we relate – navigate discussions of their sexual health and history provides a “social script” for us to mimic those behaviors in our own lives.  Basically, we’re just copying Rachel, Kurt and Jess as they navigate discussions about sex.

As extra credit for one of their courses, 234 undergraduate students watched one of three episodes of Sex in the City.  (For the record, I hope the researchers got approval from their Institutional Review Board before subjecting students to such a painful treatment condition.)  One-third of the sample watched an episode where Samantha & Co. openly discussed sexually transmitted infections (STI) and their sexual histories.  One-third of the sample watched an episode where STIs were part of the plotline, but were not openly discussed between characters.  The final third watched an episode without mention of STIs or characters’ sexual histories.  Students completed a post-test evaluation immediately after their viewing, followed by a second post-test evaluation two weeks later, detailing their own sexual behaviors and discussions.

The researchers hypothesize that television characters provide the narrative scripts for our own lives.  After watching episodes where characters discuss their sexual health and history, they argue that students are more likely to model that behavior.  Where the episodes don’t contain such discussions, students don’t learn the narrative scripts to talk about sex and sexual history.  

In the immediate post-test after watching the episodes, the research finds no difference between groups when asked about their discussions of STIs and sexual histories.  Students who viewed the episodes that included the STI storylines were no more likely to report themselves likely to discuss STIs with their partners, health care providers, or friends, and were no more likely to report that they would get tested.  However, things shifted in the two-week period following their viewing.  Students who watched the Sex and the City episode where sexual discussions took place were significantly more likely to have engaged in discussions of safe sex during the two-week period following their viewing.  In other words, “exposure to the characters modeling sexual discussions increased related behavior over the subsequent two weeks.”

The findings are particularly robust for sexually active participants in the sample.  Watching Samantha and the girls babble about STIs significantly increased the likelihood that sexually active students discussed STIs with their partners.  And while only a minority of students engaged in these discussions, “participants (who watched the storyline where STIs and sexual history were discussed) were more than twice as likely to engage in sexual health conversations after exposure (to the episode of Sex and the City) than either of the comparison conditions.”

The moral of the story is clear: More sex on TV!  When it comes to discussions of sexual history, Jess (oh god), Finn (oh, help!), Kurt (please no!) and Samantha (seriously?!) are good role models.  When they talk openly about their sexual histories, they provide a narrative script for us to talk about our own.

Worried about gaining weight during the holiday season?  Well, gorge yourself on this.  

A handful of nutritional studies suggest that the fear of holiday weight gain is (partially) for the birds.  We might gain a little weight over Thanksgiving, but we don’t leave our parents’ homes unrecognizably larger.

Research reports that people gain an average of about 0.5 kg over the Thanksgiving holiday.  (In America, that’s about equal to 1 lb, although saying you’ve gained only half a kilogram sounds better than admitting you’ve gained a pound, right?)  The sample for this study is a convenience sample of college students, so we might be cautious in generalizing across the population.  And to get their results, researchers simply weighed students before and after the Thanksgiving holiday.  Knowing they were part of a study, participants might have changed their behavior (i.e., eaten less during Thanksgiving), thereby underestimating actual weight gain over Thanksgiving.  But acknowledging these caveats, the research finds that both men and woman gained about 0.5 kg of weight during Thanksgiving.  Participants who were already overweight gained more than those who were of healthy BMI.

A follow-up study, which looked at weight gain from Thanksgiving through New Years, found no evidence of weight gain, but substantial evidence of a shift in the percentage of body fat.  While participants didn’t gain weight during the six-week period, the percentage of body fat rose and the percentage of fat-free mass fell.  

So, who’s to blame for weight gain over Thanksgiving?  The Pilgrims, of course.  After all, they were the ones who laid out lavish spreads of holiday cookies and buttery mashed potatoes before going off to their local shopping mall to buy flat screen TVs so they could sit in their cul-de-sac’ed suburbs drinking eggnog … Either the Pilgrims, or the set of cultural expectations that have developed around increasingly consumer-driven holidays.  As the public health researchers point out, “The holiday season is a time when cultural and social influences combine to create a high-risk environment conducive to weight gain.”

And speaking of high-risk environments, I’ve eaten about thirty-five cookies since arriving at my parents’ house less than twenty-four hours ago … Happy Thanksgiving. 

Over the winter, thousands of American college students will apply to Teach for America (TFA), one of the preeminent post-college teaching programs.  While the direct service provided by TFA constitutes the core goal of the program, proponents of TFA often point to the civic benefits that Corps members gain as a result of their participation.  But does Teach for America actually direct college graduates to a lifetime of civic involvement?

In a 2008 study published in Social Forces, Sociologists Doug McAdam and Cynthia Brant surveyed all the college students accepted into TFA between 1993 and 1998.  They classify their survey population into three groups – those who were accepted into TFA and completed their two-years of teaching (“graduates”); those who were accepted into TFA, but left the teaching corps before competing their two years of teaching (“drop-outs”); and those who were accepted into TFA, but declined the offer (“non-matriculants”).   They’re interested in whether graduates of TFA report stronger civic attitudes or more civic behavior after completing their service than those who were accepted into the program, but dropped out or never matriculated.   

The study reports that TFA participants are more likely to hold civic attitudes after their participation, but are less likely to engage in actual civic activities following their two-year commitment with TFA.  Graduates are more likely to support the statement, “Much of what I do is for a cause larger than myself,” or “I am willing to go to great lengths to fulfill my obligations to my country,” than both drop-outs and non-matriculants.  However, when McAdam and Brant look at actual civic participation, they find that graduates lag behind both drop-outs and non-matriculants.  Those who have completed TFA are less likely to participate in civic activities (e.g., volunteer organizations, sports leagues, etc.) and less likely to participate in institutional politics (e.g., political campaigns, party politics) that those that dropped out or never enrolled.

What explains the depressed levels of participation in civic affairs among TFA graduates compared dropouts and non-matriculants?  One possibility is service fatigue – the feeling that they’ve done their part.  An alternative is that graduates’ experiences in the classroom proved disillusioning, at least temporarily.  These negative attitudes could dissuade them from future civic commitments. The authors also raise the possibility that lower levels of civic engagement have nothing to do with TFA.  It is possible that non-matriculants found alternative service programs that increased their civic commitments more than TFA would have.  (Their survey, which asked non-matriculants what they did instead of TFA, found only 30% engaged in activist-type organizations, although the survey question is not fine-grained enough to really understand their post-college experience.) 

But McAdam and Brant really settle on a fourth answer. Among TFA graduates, they find substantial differences about how satisfied they were with their TFA experience.  When McAdam and Brant examine only graduates who were “satisfied” with their TFA experience (85 percent of graduates), they actually look no different from non-matriculants in their civic commitments.  It’s the 15 percent of dissatisfied graduates driving the difference.

Interesting as the survey findings are, they are subject to one major criticism (acknowledged by the authors).  Applicants to Teach for America (and similar programs) are already among the most civically aware group of American college students.  In this study, all of the survey respondents were accepted to Teach for America.  Both their application and acceptance signal that they are already civically engaged, and therefore asking whether service programs (like TFA) make them into better citizens may not be the most salient question.  Instead, a more interesting research design would somehow incentivize students who are not already civic-minded to participate in TFA (or a similar program) and measure the effect of these service programs on their post-program civic engagement.