Anders Holm

Comparing Regression Coefficients Between Same-sample Nested Models Using Logit and Probit: A New Method

Logit and probit models are widely used in empirical sociological research. However, the common practice of comparing the coefficients of a given variable across differently specified models fitted to the same sample does not warrant the same interpretation in logits and probits as in linear regression. Unlike linear models, the change in the coefficient of the variable of interest cannot be straightforwardly attributed to the inclusion of confounding variables. The reason for this is that the variance of the underlying latent variable is not identified and will differ between models. We refer to this as the problem of rescaling. We propose a solution that allows researchers to assess the influence of confounding relative to the influence of rescaling, and we develop a test to assess the statistical significance of confounding. A further problem in making comparisons is that, in most cases, the error distribution, and not just its variance, will differ across models. Monte Carlo analyses indicate that other methods that have been proposed for dealing with the rescaling problem can lead to mistaken inferences if the error distributions are very different. In contrast, in all scenarios studied, our approach performs as least as well as, and in some cases better than, others when faced with differences in the error distributions. We present an example of our method using data from the National Education Longitudinal Study.

Evaluating the performance of simple estimators for probit models with two dummy endogenous regressors

This study considers the small sample performance of approximate but simple two-stage estimators for probit models with two endogenous binary covariates. Monte Carlo simulations show that all the considered estimators, including the simulated maximum-likelihood (SML) estimation, of the trivariate probit model are biased in very small samples (N=100). With moderately small samples (N=500), some of the approximations perform as well as the SML estimator when the degree of endogeneity is not very large. Some of the approximations seem robust with higher correlations and are also promising for testing the exogeneity of binary covariates. The methods are used to estimate the impact of employment-based health insurance and health care (HC) on HC use, where the approximations seem to work at least as well as the SML and in some cases better.

Work for Passion or Money? Variations in Artists’ Labor Supply

Summary This paper assesses the relative impact of work for money or work for passion on Norwegian artists by examining artists’ labor supply. Our contribution is twofold. The first is to test the work-preference model and the second is to investigate the impact of arts grants on artists’ labor supply. The empirical specification draws two distinctions: between arts and non-arts income and between labor and non-labor income. Non-labor income is divided into three different sources: (1) spouses income, (2) income from financial assets and social benefits, and (3) arts grants and subsidies. Our contribution adds to the literature by estimating the significance of these various income sources on the time allocated to arts work, non-arts work, and leisure. The results provide convincing evidence for the work-preference model, and ad hoc evidence shows that art grants have a significant positive effect on the supply of arts hours. This finding supports arts policy and shows the impact of art grants on artists’ motivation to work on their arts. The causality of wages on supply is demonstrated by estimating the effects of wage shocks (grants) on arts labor supply using fixed-effect and difference-in-difference methods.

The Effect of Volunteer Work on Employability: A Study with Danish Survey and Administrative Register Data

In addition to benefiting others, volunteer work is argued to supply volunteers themselves with skills, reputation, and social connections that increase overall employability. We test this hypothesized causal link between volunteer work and employability with a high-quality 2012 Danish survey sample of 1,867 individuals of working age. The survey data are linked to administrative registers containing individual-level data on unemployment. A combination of detailed controls, lagged dependent variables, and instrumental variable regression is used to determine cause and effect. Our findings show that performing volunteer work does not statistically significantly affect the risk or rate of unemployment for the typical individual on the labour market. Introduction Unemployment constitutes one of society’s most severe social risk factors. Experiencing prolonged periods of unemployment negatively affects shortand long-term earnings (Marczak, 2009), social standing (Letkemann, 2002), and physical and psychological health (Browning and Heinesen, 2012). To avoid prolonged periods of unemployment, individuals are encouraged to invest their time in activities outside of the workplace that raise their employability (Smith, 2010). One such activity may be unpaid volunteer work, which is hypothesized to increase employability, since it may provide individuals with job-related skills; extend social networks; and signal one’s work ethic and social conscience (ibid.). Even though employability is often cited as a private benefit of volunteer work—along with the public benefits such work may bring to others—few studies have linked volunteer work to employability. Moreover, the few studies that do so rely on either anecdotal accounts or quantitative techniques that are not able to determine causal effects (Spera et al., 2015). This study fills this key gap in the research literature by empirically determining whether volunteer work has a quantifiable causal effect on risk and rate of unemployment. The study utilizes a subsample (n1⁄4 1,867) of the high-quality 2012 Danish Volunteer Survey, which contains detailed information on type and amount of volunteer work. Exact information on background characteristics and weekly unemployment status is retrieved via administrative registers. A combination of high-quality controls from the survey and administrative registers, lagged dependent variables (LDVs), and instrumental variable (IV) VC The Author 2017. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com European Sociological Review, 2017, Vol. 33, No. 3, 349–367 doi: 10.1093/esr/jcx050 Advance Access Publication Date: 29 May 2017 Original Article D ow naded rom http/academ ic.p.com /esr/article-at/33/3/349/3858046 by gest on 24 Feruary 2020 regression is used to determine the causal effect of volunteer work on risk and rate of unemployment. The empirical analysis shows that volunteers generally do not have a statistically significantly lower risk or rate of unemployment—when self-selection of highly employable individuals into volunteer work is controlled for. The robustness of this result is challenged by examining the type and intensity of volunteer work. However, this was found not to alter the overall result. Employability and Volunteer Work Individuals who are disproportionally able to achieve and maintain employment—and consequently ward off periods of unemployment—are said to have a high rate of employability (McQuaid and Lindsay, 2005). Employability is a multifaceted concept, which, according to Kanter, consists of ‘. . . a person’s accumulation of human and social capital—skills, reputation, and connections—which can be invested in new opportunities that arise inside and outside the employee’s current organization’ (Kanter, 1995). To increase one’s employability, one should therefore focus on investing time in activities that will increase one’s stock of human, social, and cultural capital (Smith, 2010). An effective way of increasing one’s stock of relevant capital resources may be by performing volunteer work. Volunteer work refers to unpaid work which benefits others outside of the home and is performed in a formal organizational setting (Wilson and Musick, 1997; Salamon and Sokolowski, 2001). Surveys show that an estimated one in every three Europeans performs such work (European Commission, 2007). The notion that volunteer work increases employability has firm intuitive and theoretical merit. Participating in volunteer work may be a way to acquire job-related skills that are advantageous when applying for work (human capital). It may also extend social networks to include individuals with information on new job positions (social capital), and finally, volunteer work may be a way to signal one’s work ethic and social conscience to potential employers (cultural capital). These mechanisms may work in tandem to increase the employability of volunteers, and it is not possible for this study to empirically highlight one mechanism over another. However, to supply a firm theoretical basis for the overall empirical investigation, each of these three mechanisms is accounted for in more detail. Human capital refers to individual resources such as knowledge and skills that enable productive activities (Schultz, 1961). It is often assumed that productive skills are required solely through education or the paid job market; however, unpaid volunteer work is likewise a productive activity that provides others with goods and services (Taylor, 2004; Wilson and Musick, 1997). In performing their unpaid volunteer duties, volunteers are, for instance, likely to be coordinating the efforts of a volunteer group, giving presentations, raising funds, or editing websites and newsletters. All of these activities can be argued to foster productive skills that are coveted on the paid job market (Wilson and Musick, 2003; Menchik and Weisbrod, 1987). Social capital refers to resources nested within social networks of two or more individuals (Coleman, 1988). Particularly social networks consisting of individuals that are not family or close friends—so-called weak ties—may be of particular importance to employability, since such networks can carry information on new paid jobs by bridging the gaps between otherwise unconnected groups of individuals, thereby linking one to opportunities nested outside of one’s immediate social circle (Granovetter, 1983). Formal organizations such as voluntary and non-profit organizations are common sources of weak ties, since they are highly likely to consist of individuals outside of one’s close-knit social circle of family and friends (Granovetter, 1973; Granovetter, 1995; Wilson and Musick, 2003; Menchik and Weisbrod, 1987; Day and Devlin, 1998). Cultural capital consists of individual tastes, attitudes, and practices that attribute to a successful presentation of the self (Smith, 2010). Volunteering is a highly regarded practice in western society, and volunteers are most often thought of in approving terms by others (Handy et al., 2010). This is likely due to the fact that unpaid work is tied to positive qualities such as personal drive, dedication to a cause, and an altruistic attitude towards one’s fellow citizens. When one takes on a role as a volunteer, it may therefore send a positive signal of one’s work ethic, moral character, and productive and social skills to potential employers (Spera et al., 2015; Handy et al., 2010). In so far as employers seek out productive and motivated employees with high levels of integrity, volunteering seems to be a good way to convey these sought-after qualities to potential employers (Smith, 2010). Time Spent and Type of Volunteer Work The underlying theoretical assumption of this study is that simply performing volunteer work may have an effect on increased employability. However, volunteer work could have a difference in treatment intensity depending on at least two factors: (i) time spent on volunteer work, and (ii) the task performed as a volunteer. Spending more time on volunteer work may enable one to take on more demanding tasks as well as increase 350 European Sociological Review, 2017, Vol. 33, No. 3 D ow naded rom http/academ ic.p.com /esr/article-at/33/3/349/3858046 by gest on 24 Feruary 2020 one’s accumulation of practical experience. Spending more time may also strengthen social ties to others within the organization and provide more opportunities to come into contact with other volunteers (Menchik and Weisbrod, 1987). Furthermore, volunteers on a board of directors are more likely to take on organizational tasks that are challenging in terms of planning and organizational skills (Balduck et al., 2009). Since there is an additional selection of high-resource individuals on a board of directors, it is furthermore likely that the value of social connections may be higher for volunteers on a board of directors (Moore and Whitt, 2000). Finally, volunteering for positions that require more responsibility, such as on a board of directors, may send a stronger signal of one’s skills and ability to future employers (Moore and Whitt, 2000). It should be considered that the above arguments hinge upon the assumption that volunteer work does indeed have a positive effect on employability. If it does not, individuals who take on more demanding tasks in terms of time and energy may actually be engaging extensively in an activity, which draws their focus away from seeking or maintaining paid employment—thus leading to a possible negative trade-off between volunteer work and paid employment (Menchik and Weisbrod, 1987). Literature Review Few studies focus specifically, as this one seeks to do, on a link between volunteer work and a quantifiable reduced rate of unemployment for the general population (Spera et al., 2015). There are studies such as the ones by Hirst (2001) and Antoni (2009) that suggest a link be

Education and Cognitive Aging: Accounting for Selection and Confounding in Linkage of Data From the Danish Registry and Survey of Health, Ageing and Retirement in Europe

Earlier studies report inconsistent associations between education and cognitive aging. We assessed the association, accounting for selective dropout due to death or dementia, and, in a subsample, accounting for confounding by early-life intelligence. Data from the Danish component of the Survey of Health, Ageing and Retirement in Europe (n = 3,400) were linked to registry data (education records, dementia diagnoses, and mortality) and the Danish Conscription Database (youth intelligence measurements for 854 men). Word recall and verbal fluency were assessed up to 4 times over 10 years (2004-2013) and combined by averaging the z scores. We fitted a joint model linking a time-to-event model for dementia or death to a linear mixed-effects model for cognitive change. Rate of cognitive decline was slower among people with high education compared with low education (β = 0.112, 95% confidence interval (CI): 0.056, 0.170). Adjusting for youth intelligence did not attenuate the association between education and cognitive decline (crude β = 0.136, 95% CI: 0.028, 0.244 vs. adjusted β = 0.145, 95% CI: 0.022, 0.269). The results suggest that higher education may slow cognitive decline in later life. In this sample, results changed little when accounting for selective attrition and confounding by intelligence.

Employment effects of active labor market programs for sick-listed workers

We use register data of 88,948 sick-listed workers in Denmark over the period 2008-2011 to investigate the effect of active labor market programs on the duration until returning to non-subsidized employment and the duration of this employment. To identify causal treatment effects, we exploit over-time variation in the use of active labor market programs in 98 job centers and time-to- event. We find that ordinary education and subsidized job training have significant positive employment effects. Subsidized job training has a large, positive effect on the transition into employment but no effect on the subsequent employment duration. In contrast, ordinary education has a positive effect on employment duration but no effect on the transition into employment. The latter effect is the result of two opposing effects, a large positive effect of having completed education and a large negative lock-in effect, with low re-employment chances during program participation.