IZA Newsroom

IZA – Institute of Labor Economics

admin

How many participants do you need to detect a treatment effect?

by

Experiments are commonly used in economics to test whether certain factors affect, for example, the behavior of consumers or the productivity of workers. When planning an experiment, several important decisions have to be made, such as how many subjects should participate. If this number is too small, there is a risk that the study cannot uncover an existing effect.

Another important question concerns the choice of design, whether participants take part in all experimental conditions (a so-called within-subjects design) or whether each participant takes part in only one experimental condition (between-subjects design). Both decisions imply tradeoffs. Between-subjects designs require more participants than within-subjects designs. But it is generally not known how much more participants are needed.

In a new IZA Discussion paper, Charles Bellemare, Luc Bissonnette and Sabine Kröger demonstrate how ex-ante power calculations can be conducted in a flexible way using simulations in order to compute the minimal number of participants. They show the trade-off between the choice of design and minimal number of participants needed. In their application of labor market gift giving field experiments, they find that between-subjects designs require 4 to 8 times more participants than within-subjects designs to reach the same power to detect an effect. The authors offer a STATA software package that runs the simulations and can be used to compute the minimal number of sample size needed to detect an effect for a variety of situations for different design choices.

Online quizzes can motivate students to learn more

by

Online QuizA large fraction of students fail at university. One potential reason is that students do not exercise enough effort on a regular basis. If intrinsic motivation does not suffice to induce satisfactory student performance, then what interventions might help to increase student effort and performance?

In a recent IZA Discussion Paper, Arnaud Chevalier, Peter Dolton and Melanie Lührmann vary incentives for students to provide effort on a weekly basis. They focus only on one type of effort, participation to a weekly online quiz which provides students with feedback of their understanding of the lecture.

On a given week, students face either no incentive, get additional educational material if they participate, the best performer wins a book voucher, or the quiz is declared to be compulsory. In a second cohort, two additional incentives are included, the quiz grade counts for 2.5% or 5% towards the final grade for the course.

The study finds that the provision of additional educational material has little impact on weekly effort, whereas the book voucher rewarding only the top performer even reduces participation. But if effort is rewarded in terms of grades, then participation becomes close to what it is under compulsion. Assessment weighting increases quiz effort and continuous learning relatively more among lower ability students.

For the cohort subjects to the assessment weighting of quiz grades final grades improve at an average of 4%. These performance increases are in the order of magnitude of the results for large financial incentives. Since all incentives in the setup relate directly to course outcomes and are easy to scale up at a low cost, the authors conclude that it is quite easy to increase students effort and grades.

Read the abstract or download the complete discussion paper [PDF].

Beware of moral hazard: Joint unemployment insurance for the EMU

by

The ongoing crisis calls for better fiscal cooperation within the European Economic and Monetary Union (EMU). Beyond tax policy coordination, there is also a strong demand for shock-absorbing mechanisms that could replace the lost instrument of exchange rate adjustment. The introduction of an EMU-wide unemployment insurance mechanism is currently debated as a potential tool to stabilize output and employment in recessions.

A new IZA Policy Paper by Werner Eichhorst and Florian Wozny analyzes and compares two different proposals:

  1. A basic EU-wide unemployment insurance which uses direct contributions from employers and employees to provide a permanent minimum level of insurance benefits to the unemployed. Countries could increase benefit generosity through national contributions.
  2. A “kicking-in” scheme that would supplement or extend national unemployment insurance schemes whenever predefined indicators, such as the interest rate of the unemployment rate, exceed a certain threshold.

Such insurance instruments naturally rely on heterogeneous developments of incurred risks. If insured countries were hit by a homogenous external shock, the system would be in trouble because all countries would need possible transfers for themselves. Although economic trends trend to converge between countries with extensive trade links, the recent crisis in the EMU showed that there would have been ample room to maneuver. At least in the recent past, a number of countries with relatively robust economies could have served as net contributors, which is the necessary condition for an insurance mechanism to work.

However, given that both EMU-wide unemployment insurance schemes are potentially vulnerable to moral hazard and manipulation, the design is of main interest. The existing proposals do not capture the main motivation of such a scheme in their design, which is to combat credit market constraints. Bank lending constraints, which led to high interest rates in the recent crisis, are a key argument for a transfer mechanism in the form of a transnational automatic stabilizer. Countries like Greece, Spain, Portugal and Ireland were unable to stabilize out-of-work income maintenance and support for the unemployed. Particularly in strong recessions, therefore, transnational unemployment insurance transfers would ease budget constraints in the hardest-hit countries.

A kicking-in scheme, linked to interest rate thresholds, captures these constraints. It would also reduce the risks of moral hazard and manipulation compared to a basic scheme. If policymakers actively tried to cross those threshold values in order to become entitled to transfers, this would be associated with the direct costs of higher interest payments, which makes manipulation less tempting. A further reduction of moral hazard could be achieved by linking transfers to structural reforms and other threshold values like the unemployment rate. Overall, this makes a kicking-in scheme appear superior to a basic scheme.

Should housework be shared more equally between partners?

by

Women do more housework than men. For most of history this phenomenon went largely unremarked. Housework was what women did, while men supported their families with paid work. However, the role of women has been changing: women’s educational levels are now at least as good as those of men; most women do paid work for most of their adult lives; and although gender pay gaps still exist, women’s earnings (in particular, the earnings of younger childless women) have been gradually approaching those of men. While things have also been changing in the domestic sphere, with men doing an increasing share of housework and childcare, surveys still show that women do the lion’s share of housework – even in households where both partners have full-time paid jobs.

Several theories have been proposed as to why unequal shares of housework have persisted even in the context when women do about as much paid work as men. The full range of these theories is discussed in a new IZA paper. The work by Katrin Auspurg, Maria Iacovou and Cheti Nicoletti investigates one strand of theory, namely the suggestion that men and women have systematically different preferences over housework because of internalized gender norms. Variants of this theory, which essentially proposes that women like housework more (or dislike it less) than men do, have been put forward by both economists and sociologists.

Previous attempts to test this theory empirically have been based on information on reported satisfaction or well-being gathered from surveys. These potentially suffer from a number of problems. First, surveys find so few households in which the man does most of the housework that it’s impossible to obtain reliable estimates of how people feel about these arrangements. Second, there is a potential problem of post-hoc rationalization – it’s likely that reported preferences are affected by people making the best of the situation in which they actually find themselves. And third, because preferences may be affected by some of the same factors which drive the amount of housework that people do, it’s almost impossible to work out the direction of any causal relationships.

The new paper takes a different approach, using data from an experiment conducted in the course of the Innovation Panel of Understanding Society. Men and women were presented with hypothetical scenarios (“vignettes”) in which several factors varied: the distribution of housework; the distribution of paid work; earnings; the presence of children; and whether the couple had paid help with housework. Each respondent was presented with three different scenarios, and was asked to rate how satisfied they would be with each scenario on a scale from 1 (completely dissatisfied) to 7 (completely satisfied). In all, 4,547 valid responses were obtained from 1,609 respondents.

The results show that:

(1) Both men and women prefer scenarios in which housework is shared more or less equally between members of a couple. They dislike scenarios in which their partner does most of the housework almost as much as they dislike scenarios in which they themselves do most of the housework.

(2) People consider paid work and housework in combination when assessing the scenarios. The general unpopularity of unequal housework distributions is reduced in scenarios where the partner doing more housework is doing less paid work, and increased in scenarios where the partner doing more housework is doing more paid work.

(3) Men’s and women’s preferences are remarkably similar over virtually the entire range of possible scenarios. The only discernible difference between the sexes is that women are less favorably disposed than men to working full-time themselves while their partner works part-time. But this difference is small and relates to paid work rather than housework.

The authors therefore conclude that there is no evidence that preferences over housework differ between the sexes as the result of internalized gender norms, and that the reasons behind the gendered allocation of housework must lie elsewhere, perhaps in the different bargaining strategies employed by men and women.

Can migration boost incomes, happiness and freedom satisfaction?

by

Carol Graham, Milena Nikolova

By Carol Graham and Milena Nikolova

The recent economic crisis and the subsequent lagging economic recovery renewed the immigration debate in the United States and beyond. While there are benefits from migration for the recipient economies, there are also potential costs, which range from concerns about migrants adding to the ranks of welfare systems, to threats to the jobs of natives, to the difficulties of assimilating migrants from different countries and cultures. On balance, migration’s benefits for the destination economies outweigh the costs: immigrants are net contributors to public coffers and complementarities between low-skilled immigrants and natives may exist, particularly when new migrants take low-skilled jobs that native workers eschew. While the effect of migration on the well-being of native populations is important, so is the question of how migrants fare once they reach their destination countries.

In part, this is because of the sheer magnitude of migration stocks: about 232 million people lived outside their country of birth in 2013. Most migrants move across international borders to maximize their earnings and to gain opportunities, and most do make significant income gains. Whether these earnings gains are mirrored by improvements in reported well-being and broader quality of life is still largely an open question, however.

Higher levels of immigrant well-being can be instrumentally important for social outcomes such as public health and productivity, as happier individuals are typically more productive and healthier. In contrast, immigrant dissatisfaction may reflect lack of assimilation and social exclusion, and even extremist attitudes among natives, all of which can result in social unrest and lower economic output. From the point of view of the sending countries, meanwhile, emigrant well-being is important as migrants send remittances and contribute to the well-being of their home countries through investments, the spread of ideas, and technology.

Understanding the well-being consequences of migrating is challenging, as comparing migrants with those who did not migrate in either the sending or the destination countries is methodologically flawed. Such comparisons may simply reflect the traits of those who choose to migrate, and who may have differences in ability, risk tolerance, aspirations, and motivation, for example. The direction of causality between well-being and migration is also unclear: while migration may influence well-being, the reverse can also be true. While migration may indeed affect happiness, a recent paper using South American data finds that respondents who intend to migrate in the next year are wealthier and more educated than the average, but also less happy and more critical of their current and future economic opportunities.

Despite the large income gains typically associated with migrating, it can also be accompanied by declining happiness because of adaptation and rising aspirations. While migrants’ (absolute) incomes increase, so do their expectations as they compare themselves to high-earning natives in the host countries. In a recent IZA Discussion Paper, we use Gallup World Poll data and statistical techniques to understand the well-being consequences of migration for movers from transition economies to advanced nations. Like other studies, we find unequivocal income increases due to migration. More importantly, we show that there are significant gains in life satisfaction and in perceptions of freedom. (See Figure 1 for the average well-being outcomes of migrants and stayers in the pre- and post-periods).

We studied migrants from transition and post-transition societies as they are quantitatively the most important migration source for the OECD economies in Europe. To illustrate, in 2012, Poland and Romania were among the top three migrant sources of OECD migrants. Migrants from transition economies leave countries, which are relatively more advanced and culturally similar to the destination countries, compared with movers from Asia, Africa, and Latin America, for example.

We find that migrants from transition countries achieve a better quality of life after they go abroad. The average household income premium from migration for our sample is about 21,000 international dollars (about 10,500 ID per household member). And even if reference norms and aspirations change, migrants’ life satisfaction also improves. The average benefit is substantively and statistically significant: an increase of about 1.0-1.2 (on a life satisfaction scale 0-10). Third, and perhaps most important, migration positively affects perceptions of freedom of choice, with migrants from the most recent EU enlargements being nearly 40 percent more satisfied with their freedom. Data from around the world show that freedom to seek life fulfillment is a pivotal element of human well-being.

At a time when there is ample reason to be concerned about the state of world affairs, our research findings are a “happy” story highlighting that by voting with their feet, migrants from transition economies can improve their well-being. Surely migration does not solve the problems in the countries the migrants left, nor can we be confident at this stage that our results apply to migrants from different origins and/or to those migrating to different destinations. Yet it does suggest that win-win outcomes are possible from increasing mobility in global labor markets. That is a story worth learning more about.

This post also appeared on the Brookings Institutions’s UP FRONT blog.

Milena Nikolova is a Research Associate at IZA and a Nonresident Fellow at Brookings. Carol Graham is the Leo Pasvolsky Senior Fellow at Brookings, College Park Professor at the University of Maryland, and a Research Fellow at IZA.

Why aging and working makes us happy – in four charts

by

Carol Graham, Milena Nikolova

By Carol Graham and Milena Nikolova

In the past few years, economists and other social scientists have made great strides in developing measures to assess subjective well-being (or, more colloquially, happiness), which has deepened our understanding of well-being beyond the traditional income dimensions. There are remarkably consistent patterns in the determinants of subjective well-being across people within and across countries and cultures around the world. One of the most striking of these is the relationship between age and happiness (which is good news for those of us who are already on the “back-nine”). There is a U-shaped curve, with the low point in happiness being at roughly age 40 around the world, with some modest differences across countries. It seems that our veneration of (or for some of us, nostalgia, for) youth as the happiest times of our lives is overblown, the middle age years are, well, as expected, and then things get better as we age, as long as we are reasonably healthy (age-adjusted) and in a stable partnership.

There are other consistent patterns. Income matters to individual happiness in every context we have studied this relationship. Yet after basic needs are met, other things like how your income compares to that of your peers also start to matter. Moreover, married people (and those in a civil union) are typically happier than their non-married counterparts (there is a direction-of-causality-issue here, though, as happier people are more likely to marry each other); healthier people are happier; and women are, in most places, happier than men (as long as gender rights are not severely compromised).

Another variable that is absolutely critical for subjective well-being is employment status. The unemployed are less happy than the employed worldwide. And both psychologists and economists find that long-term unemployment has psychological scarring effects. Long-term unemployment and under-employment, and the youth’s delayed entrance into employment, coupled with the over-burdened pension systems, are major problems in the U.S. and Europe. At a time when these issues have risen to the fore, it is perhaps worth considering more flexible labor market arrangements. While several solutions have been proposed, we are left wondering whether there would be public receptivity to changing labor market arrangements. While this is hard to predict, what we can measure – and did in our recent article in the IZA Journal of European Labor Studies – is the well-being costs and benefits of different work arrangements. As we argue in the study, different employment and retirement arrangements may be appropriate for people at different stages of their lives, depending on their career goals or innate well-being levels. Understanding how employment, retirement, and late-life work relate to well-being can contribute to ongoing public policy discussions. [Read more…] about Why aging and working makes us happy – in four charts

Immigrant women: Not just secondary workers any more

by

Immigrant women in the labor market have long been viewed as “secondary workers” who work in unskilled jobs, mainly as a response to family needs and to support their husbands’ skills upgrading. As household financial constraints ease with men’s assimilation in their destination country, women’s labor market participation is expected to drop again. An IZA Discussion paper by Alícia Adserà and Ana Ferrer on immigrant women in Canada finds that this is no longer true.

The authors combine data of over 800,000 women in Canada for the period 1991 to 2006 with information on the skill requirements of the jobs women hold (such as physical strength or analytical abilities) to examine the labor market patterns of immigrant women. They show that the recent behavior of married immigrant women does not fit the profile of secondary workers, but rather conforms to the recent patterns of native Canadian wives, with rising participation in the labor market and wage gains as they stay longer in Canada.

Immigrant women’s behavior displays a path of assimilation similar to that of immigrant men. They increasingly make labor supply decisions guided by their own opportunities in the labor market rather than by their husbands’ trajectories. At best, only uneducated immigrant women in jobs requiring very basic skills may fit the profile of secondary workers with slow skill mobility and low-status job-traps.

Solomon Polachek on equal pay and the gender wage gap

by

In a video interview, IZA Fellow Solomon Polachek (State University of New York at Binghamton) talks about equal pay and the gender wage gap.

He explains that equal pay policies based on wage outcomes have had little effect on the gender wage gap. Instead, economic policies that promote human capital accumulation and greater lifetime work for women – e.g. subsidizing day care and abolishing marriage taxes – can successfully reduce the gender wage gap further.


 

For more information on equal pay and the gender wage gap see a related article by Solomon Polachek entitled Equal pay legislation and the gender wage gap, which has been published on IZA World of Labor.

Where does the gender gap in economics enrolment arise?

by

Although women account for 57% of all students at UK universities, the share of female economics students is only about 27%. In a new IZA paper, Mirco Tonin and Jackie Wahba use rich data on university applications to explore the reasons behind the big gender gap in economics enrolment. They also explain why it would be in the interest of society to close this gap.

Economists generally have an influential role in policy making, directly or as academics, consultants or policy analysts. The gender of those making policies matters because males and females tend to have different policy preferences. Also, since economics is one of the subjects associated with relatively high average earnings, lower enrolment by females is a contributing factor behind the gender pay gap.

The study finds no indication that universities discriminate against females when making their offers. As the acceptance rate does not differ between males and females, the enrolment gap is entirely due to low application rates: Only 1.2% of females apply to economics, as opposed to 3.8% of males. An important limiting factor (though not the whole story) seems to be the gender gap in math. Among those who enroll at university, math is an A level subject for 19% of males, but only 10% of females.

There is evidence that this gap in math is cultural rather than “biological”, as the gap disappears in more gender-equal societies. Thus, a policy aimed at reducing the gap in math could be very effective. In particular, it could start a positive loop, with an improvement in the participation of girls to educational paths leading to positions of influence generating more equality in society, again encouraging girls to seek educational paths leading to positions of influence, and so on.

Grades and rank: How do they affect student effort?

by

The educational system is built upon a sequence of tests to measure student performance. Improving student performance is a key educational and economic policy issue to which much time and resources are devoted. At the same time, the results of these tests often lay the foundation for the distribution of school resources and spur considerable public debates. A question that has not received as much consideration is how test incentives inherent in our conventional grading system affect student motivation and effort? What if we leave the beaten path of grading students on scale from 1 to 6 or A to F and introduce alternative ideas of incentivizing students?

In a new IZA Discussion paper Nina Jalava, Juanna Schrøter Joensen and Elin Pellas conduct a randomized field experiment on more than a thousand sixth graders in Swedish primary schools to shed light on the effects of alternative incentive systems in the classroom. They find that ranking students by distinguishing the top three performers in the class has particularly large motivational power for highly-skilled boys and girls. While the boys seem to compete independent of the peer group, highly skilled girls seem less motivated if they are tested in a class where they do not know their classmates well. In addition, these girls also increase their effort if offered a certificate for high performance – an incentive boys do not react to. However, looking at the low skilled students, these incentive measures may backfire and exert negative effects on effort.

Based on these results, the author call for caution in using financial incentive schemes for schools and teachers – such as school accountability or teacher performance pay systems — based on the outcomes of students’ test scores. The study implies that the distribution of financial resources will be highly dependent on how the students prepare for the tests, whether students know their peers during the tests and how teachers set incentives in terms of grading. It could therefore be easy for teachers and principals to manipulate such measures by simply giving a symbolic reward or singling out the top performers on the day of the test, which would in turn mask and bias true student ability.