Gender Gaps Conference 2021

Sześć podstawowych pytań badawczych naszego projektu.

1. Czy nierówności na rynku pracy sprzyjają podejmowaniu samozatrudnienia z przymusu?
   Analizujemy kanały, przez które nierówności na rynku pracy mogą wpływać na decyzję o samozatrudnieniu, formułujemy model teoretyczny i przeprowadzamy empiryczną weryfikację jego przewidywań, dostarczając w ten sposób nowych wniosków na temat powiązań między nierównościami a samozatrudnieniem.
2. Czy start-upy kierowane przez kobiety podlegają silniejszym ograniczeniom zasobów niż start-upy założone przez mężczyzn? Czy wpływa na to dalszy rozwój firmy?
   Dotychczasowe badania wykazały, że przy tworzeniu startupów występuje tzw. dobór assortatywny. Korzystamy z bogatych danych dopasowanych do pracodawców i pracowników, które obejmują kilka dekad wstecz dla Norwegii, aby zbadać uprzedzenia płciowe w nowych firmach venture capital.
3. Czy zróżnicowanie rad nadzorczych pomaga zwiększyć zróżnicowanie najwyższego kierownictwa, zwłaszcza wśród firm niepodlegających kontroli publicznej?
   Wykorzystując unikalne i kompleksowe dane dla ponad 100 milionów firm w Europie, a także innowacyjne metody data science badamy przypadki wejścia kobiet do rad nadzorczych i obserwujemy ich wpływ na zróżnicowanie na pozycjach zarządczych w tych firmach.
4. Czy różnorodność rad nadzorczych i zarządów ze względu na płeć wpływa na wyniki firmy?
   Przedstawiamy pierwsze przyczynowe i kompleksowe szacunki dotyczące wpływu różnorodności w radach płci na wyniki firmy. Nasze badanie byłoby kompleksowe, ponieważ zbada skutki nie tylko dla firm publicznych (notowanych na giełdzie), ale także dla firm prywatnych (nie notowanych na giełdzie). Będzie również kompleksowy w tym sensie, że nasza strategia identyfikacji opiera się na zwykłym biznesie, a nie na konkretnym, zazwyczaj nietypowym czasie reformy przepisów dotyczących ładu korporacyjnego. Wkładem naszej pracy jest połączenie metodologii tzw. instrumentów Bartika z weryfikacją kryterium wykluczenia w data science (uczenie maszynowe) w celu zapewnienia kompleksowej oceny wpływu zróżnicowania płci na wyniki firmy.
5. Jaka jest rola unikania informacji i potwierdzania stronniczości w ocenie wyników menedżerów mniejszościowych w porównaniu z menedżerami spoza mniejszości?
   Osadzone w ramach unikania informacji, zaprojektowaliśmy dwa eksperymentalne ustawienia wyróżniające się interesem własnym uczestników (w jednym projekcie badani zaobserwują bezpośrednią korzyść finansową ze swoich decyzji, w drugim projekt odniesie korzyść dla społeczeństwa). Wnioskujemy o powiązaniach między postrzeganiem poszczególnych umiejętności i zdolności w zależności od płci a wpływem kontekstu na te postrzeganie.
6. Jaka jest rola confirmation bias w ocenie potencjalnych wyników kandydatów do pracy z mniejszości?
   Eksperymentalnie weryfikujemy występowanie potencjalnie niepożądanych konsekwencji wiążących klauzul równości szans.

The key six research questions behind our project are summarized as follows.

1. Does labor market inequality drive take-up of necessity self-employment?
   Theorizing about the channels through which labor market inequality may influence the choice to become self-employed, we develop a theoretical setup and provide empirical verification of its predictions, thus yielding novel insights into the links between inequality and self-employment.
2. Are minority-led start-ups subjected to stronger resources constraints than start-ups established by non-minority entrepreneurs? Does it affect subsequent firm growth?
   Existing research has demonstrated that there occurs assortative matching in the formation of founding teams, albeit with many methodological caveats and theory shortages. We exploit rich employer-employee matched data that cover several decades back for Norway in order to investigate gender biases in new venture firms.
3. Does diversity of the supervisory boards help to increase diversity of top management, especially among firms who are not subject to public scrutiny?
   Using unique and comprehensive data for 100+ million firms in Europe, as well as innovative data science methods, we study the events of women entering the supervisory (non-executive) board and observe the subsequent changes (or lack thereof) in the composition of top management.
4. Does diversity of the supervisory and management boards affect firm performance?
   We provide the first causal and comprehensive estimates for the effects of gender board diversity on firm performance. Our study would be comprehensive in that it will study the effects not only for the publicly held firms (stock listed) but also for the privately held firms (not listed on the stock exchange). It will also be comprehensive in a sense that our identification strategy is based on business-as usual rather than on specific, typically unusual timing of a corporate governance regulations reform. A contribution of our work is to combine methodology of the so-called Bartik instruments with data science (machine learning) verification of the exclusion criterion in order to provide a comprehensive evaluation of the effects of gender board diversity on firm performance.
5. What is the role of information avoidance and bias confirmation in evaluating performance of minority managers when compared to non-minority managers?
   Embedded in information avoidance framework , we designed two experimental settings distinguished by self-interest of participants (in one design the subjects will observe a direct financial benefit from their decisions, in the other design the benefit will be to the society). We infer the links between the perceptions of individual skills and abilities across genders and how these perceptions are affected by context.
6. What is the role of confirmation bias in evaluating potential performance of minority job applicants?
   We experimentally verify prevalence of potentially undesirable consequences of binding equal opportunity clauses.

Goldin (2014, AER) argues that societies have made a remarkable progress in closing the wage and employment gaps in the last fifty years; yet, they have failed to achieve full labor market equality. Goldin boils down the remaining difference to the concept of time flexibility, which should be understood both from employee and the employer perspectives. For example, primary care givers have lower time endowment to allocate to market work. By the same token, groups disadvantaged due to remote location, disability etc. may be at disadvantage when compared to workers without such handicaps. While the conjecture of Goldin is attractive, empirical evidence is scarce and at times contradictory. In this project, we provide a battery of tests Goldin’s conjecture, analyzing the role of working time flexibility in determining wages and employment.

First, from an employee perspective, we hypothesize that demand for flexibility varies at different stages of the life-cycle. If the demand for flexibility indeed drives the (adjusted) wage gaps, then one expect a life-cycle pattern in adjusted wage gaps. In the absence of flexible arrangements, the demand for flexibility might also be reflected in more selective employment patterns. This hypothesis will be tested empirically, with the use of a novel proposed estimation method, which allows to isolate age, cohort and time effects in adjusted wage gaps. The study will be done for a wide variety of countries. Second, also taking the employee perspective, we will infer the true value of working time flexibility to workers. Non-standard work arrangements allow to engage in other spheres of life, but in many societies carry a social stigma. We will propose a large-scale framed field experiment to infer the true value of working time flexibility among workers, controlling for their outside options and household situation.

Third, from an employer perspective, we formulate the hypothesis that managers from a disadvantaged group will not tolerate high wage penalty on time inflexibility among their subordinates. To this end, we will develop and utilize a novel dataset on female managers across Europe and industry-level adjusted gender wage gaps (as well as their between firm dispersion). We will provide a variety of identification strategies to inspect the causality in this relationship.

Fourth, from a household perspective, taxes and social transfers may disproportionally tax the second earner’s labor supply, thus making it disadvantageous to supply labor. So may the childcare costs. We will extend a standard tax-benefit microsimulation model to account for the latter for a selection of the European countries, thus obtaining a shadow price of working time flexibility across countries. Finally, fifth, we will analyze the links between the supply of flexible work agreements and labor market inequality. We will explore ways flexible working arrangements influence labor market inequality across the EU, including selective patterns of employment, changes in within gender inequality and assortative mating.

The project contributes to the literature in three respects. First, we provide a wide battery of tests to an important hypothesis concerning the origins of labor market inequality in a wide, comparative context across countries. Second, we provide several methodological innovations (concerning estimation techniques and identification strategies) as well as methodological diversity (econometrics, simulations and experiments). Third, we will separate the role of the supply side factors (on the side of the employers) from demand side factors (the working time flexibility to combine personal and professional life).

The benefits of international cooperation are threefold. First, the teams complement each other in terms of skills, which permits pushing the frontier of research. Second, given the international composition of the team, the comparative economics angle can be based on local expertise on top of being data driven. Third, the capacity building for both partners will permit broader reach of this and future research in the field, thus contributing to the internationalization of Polish and Lithuanian science in our field.