W statystyce łatwo znaleźć jest odpowiedź, wynik analizy, ale znacznie trudniej o pewność, na które pytanie jest to odpowiedź.
Hubert
Drążkowski
Hubert joined us to work on EARHART team. He has obtained a Master's degree at faculty of Mathematics and Information Science in mathematical statistics and data science. More bio on https://www.linkedin.com/in/hubert-dr%C4%85%C5%BCkowski/
-
Czy możemy kontrolować zagrożenia AI? Sprawdza Hubert Drążkowski GRAPE | Tłoczone z danych dla...
-
Jak mierzyć ludzki potencjał? Sprawdza Hubert Drążkowski GRAPE | Tłoczone z danych dla...
-
Czy algorytmowi udałoby się przewidzieć kolejne wydarzenie w życiu człowieka na podstawie jego historii? Sprawdza Hubert Drążkowski GRAPE | Tłoczone z danych dla...
-
Jak statystyka pozwala odkrywać niewidzialne bariery? Sprawdza Hubert Drążkowski GRAPE | Tłoczone z danych dla...
-
Czy kosmiczna ekonomia będzie wyglądała tak samo jak na Ziemi? Sprawdza Hubert Drążkowski GRAPE | Tłoczone z danych dla...
-
Węże i ośmiornice – jak rysowanie mapy okręgów wyborczych wpływa na ostateczny wynik głosowania? Sprawdza Hubert Drążkowski GRAPE | Tłoczone z danych dla...
-
Czy kosmici są wśród nas? Sprawdza Hubert Drążkowski GRAPE | Tłoczone z danych dla...
-
Projekt Huberta Drążkowskiego "Sponsorowane treści w ramach modelu wielorękiego bandyty"...
-
Czy życie to bardziej złożone klocki lego? Instrukcji budowania szuka Hubert Drążkowski GRAPE|Tłoczone z danych dla...
-
Udział płac w wartości dodanej w Polsce nie spada. Od dekady utrzymuje się na niemal niezmienionym poziomie i wynosi 57% – wynika z najnowszego badania Sebastiana Zalasa i...
-
Aaaaby język giętki, powiedział wszystko, co pomyśli głowa… A co jeśli ma zagmatwać tak, aby głowa nie złapała? Hubert Drążkowski o żargonie prawniczym i...
-
Jak sztuczna inteligencja pomaga przewidywać pogodę? Spoiler: wyniki są bardzo obiecujące, a to dopiero początek! Felieton Huberta Drążkowskiego GRAPE|Tłoczone z danych...
-
Wiesz jak myśli sztuczna inteligencja? Naukowcy też nie (do końca) wiedzą. Postanowili to zmienić. Ich wysiłkom przyjrzał się Hubert Drążkowski w GRAPE|Tłoczone z danych...
-
Nie dla wszystkich los łaskawy... Na szczęście z pechem można walczyc. GRAPE|Tłoczone z danych...
-
-
Opublikowane | Published
-
Gender board diversity across Europe throughout four decades | Nature Scientific Data Przeczytaj streszczenie | Read abstract
We present a Gender Board Diversity Dataset (GBDD), which provides a cross-country perspective on women in management and supervisory boards that spans between 1985 and 2020. The data covers 43 European countries and accounts for private companies in addition to the stock-listed ones. GBBD was created using firm-level Orbis data. Our measures are based on a sample of more than 28 million unique firms observed for nearly seven years on average and reporting data about nearly 59 million individuals on management and supervisory boards. We provide the measures at the level of industry, country and year (the firm-level data is proprietary). We provide three measures. The first is the share of women among all board members in a given industry, country, and year. The second one is the average of the shares of women across firms in a given industry, country and year. We also provide a new measure: the share of firms in a given industry, country and year which report no single woman on their board(s).
-
The evolution of labor share in Poland. New evidence from firm-level data | Gospodarka Narodowa Przeczytaj streszczenie | Read abstract
The only accessible source of financial firm-level in Poland data is database prepared by Central Statistical Office (CSO), available for only minor group of economists. Our study aim at providing to wider audience Orbis database, providing information about substantial part of Polish firms. With this data we present the evolution of labor share of value added in Poland, together with labor productivity and wages. We compare our estimates with indicators prepared by Growiec (2009) and national accounts data. We find that different data source produce differing levels of indicators, however its shapes remain similar. Furthermore we show how construct database using Orbis data and how to improve its coverage using imputation methods.
W toku | Work in progress
-
Gender tokenism in corporate boardrooms in Europe Przeczytaj streszczenie | Read abstract
While women constitute 25% of the board members in private European corporations, firms with no women are dominant and merely 10% of firms report more than one woman in the boardroom. We leverage new data, covering 5 million companies from 29 European countries spanning 1986-2020. Using this data, we provide novel stylized facts about the phenomenon of token women in European corporate boardrooms. We emphasize the differences between two types of situation when the corporate boardroom hosts a singular woman: the appointment of the first woman and the barriers to further rise in diversity. We thus contribute to improving our understanding of the conflict between the tokenism hypothesis and diversity spillover hypothesis.
-
Gender board diversity spillovers and the public eye Przeczytaj streszczenie | Read abstract
A range of policy recommendations mandating gender board quotas is based on the idea that "women help women". We analyze potential gender diversity spillovers from supervisory to top managerial positions over three decades in Europe. Contrary to previous studies which worked with stock listed firms or were region locked, we use a large data base of circa 2 000 000 firms. We find evidence that women do not help women in corporate Europe, unless the firm is stock listed. Only within public firms, going from no woman to at least one woman on supervisory position is associated with a 10-15% higher probability of appointing at least one woman to the executive position. This pattern aligns with signaling, stakeholder and institutional theories, suggesting that external visibility influences corporate gender diversity practices. The study implies that diversity policies, while impactfull in public firms, might have limited effectiveness in promoting gender diversity in corporate Europe.
-
Implicit gender quota in European boardrooms Przeczytaj streszczenie | Read abstract
We test for implicit gender quotas in the boardroom. We use novel dataset covering 11 million European corporations over three decades. We find that -- accounting for the pool of available candidates -- gender-blind hiring of women to board positions is highly improbable. Implicit quotas refer to unspoken policies or practices that result in a specific gender composition. Tokenism is one such example: in order to project the reputation of supporting diversity, an organization may prefer to invite a single representative of minority (or representatives of minorities) without endowing them with actual decision power.