Dokument: Four Essays in Empirical Trade and Labor Economics
| Titel: | Four Essays in Empirical Trade and Labor Economics | |||||||
| URL für Lesezeichen: | https://docserv.uni-duesseldorf.de/servlets/DocumentServlet?id=50428 | |||||||
| URN (NBN): | urn:nbn:de:hbz:061-20190814-090418-0 | |||||||
| Kollektion: | Dissertationen | |||||||
| Sprache: | Englisch | |||||||
| Dokumententyp: | Wissenschaftliche Abschlussarbeiten » Dissertation | |||||||
| Medientyp: | Text | |||||||
| Autor: | Knauth, Florian [Autor] | |||||||
| Dateien: |
| |||||||
| Beitragende: | Prof. Dr. Südekum, Jens [Gutachter] JProf. Dr. Dauth, Wolfgang [Gutachter] | |||||||
| Stichwörter: | Internationaler Handel, Arbeitsmarktökonomik, Migration, Ungleichheit | |||||||
| Dewey Dezimal-Klassifikation: | 300 Sozialwissenschaften, Soziologie » 330 Wirtschaft | |||||||
| Beschreibung: | The access to high quality individual data has drastically increased in the last 20 years. The possibility to observe individuals across time, space and employer has a
high value for researchers, both for describing the status quo and producing stylized facts about the economy as well as designing research studies and identifying causal effects. It can help understand major political events and subsequently to give informed policy advice. The abundant information has to be structured and reduced to be usable. In my work I focus on two things. Firstly, I make use of a large individual panel data set and retrieve additional information on the workers by observing their behavior and realized labor market variables, such as wages, occupations and unemployment information. In particular, I use the network structure of the data to estimate latent characteristics of workers and firms. The use of large and almost full samples of the workforce allows to observe a tight network between them. Secondly, I develop empirical hypotheses and connect them to economic theory. In particular I am interested in empirical studies on the relation of trade, technological change and wage inequality in Germany as well as the reasons for internal migration in Germany. For this purpose I connect the information on workers and firms to other sources of data. In particular, information on global trade flows and automation of production, which are main drivers of wage inequality. In our study in the first chapter we use a large sample of German workers to analyze the effect of low-wage competition with China and Eastern Europe (the East) on the wage structure within German manufacturing industries. Utilizing the method by Abowd, Kramarz, and Margolis (1999) (hereafter AKM), we decompose wages into firm and worker components. We find that the rise of market access and competitiveness of the East has a substantial impact on the dispersion of the worker wage component and in part on positive assortative matching. Trade fails to explain changes in the firm wage premium. The rising dispersion in worker-specific wages can be attributed to increasing skill premia and to changes in the extensive margin of the workforce, leading to a wage polarization for the remaining within-industry workers. We also account for technological change by considering how many routine-intensive jobs are substituted within an industry. The more routine jobs are cut, the higher is the effect on wage inequality, especially on the dispersion of worker-specific wages. Overall, trade explains up to 19% of the recent increase in wage inequality and slightly exceeds the technology effect that accounts for approximately 17%. In chapter 2 we present supporting empirical evidence and a new theoretical explanation for the negative selection into planned return migration between similar regions in Germany. In our model costly temporary and permanent migration are used as imperfect signals to indicate workers’ high but otherwise unobservable skills. Production thereby takes place in teams with individual skills as strategic complements. Wages therefore are determined by team performance and not by individual skill, which is why migration inflicts a wage loss on all workers, who expect the quality of their co-workers to decline. In order to internalise this negative migration externality, which leads to sub-optimally high levels of temporary and permanent migration in a laissez-faire equilibrium, we propose a mix of two policy instruments, which reduce initial outmigration while at the same time inducing later return migration. While we show a theoretical explanation for internal migration in chapter 2, in chapter 3 we use a large German administrative dataset and assess sorting patterns of workers between high-density regions in Germany. With detailed wage information we predict the selection of workers into well defined mobility patterns such as permanent, return and move-on migration. We assess latent heterogeneity in the wage data by estimating premia to unobservable skill for workers and firm pay premia (Card, Heining, and Kline, 2013, hereafter CHK). We show that being matched to a low-paying firm strongly increases the probability to return migrate. In terms of unobservable skill we find that initially migrants are positively selected. Given migration occurred, return migrants are negatively and move-on migrants are highly positive selected. Move-on migrants benefit from additional moves both in terms of skill and firm premium. Finally, in the fourth paper we again turn to a large random sample of the German workforce to assess sorting patterns in a matched employer-employee framework. We compare different econometric models to estimate bipartite latent heterogeneity of workers and firms using recent methods proposed by Card, Heining, and Kline (2013) and Bonhomme, Lamadon, and Manresa (2017) (hereafter BLM). With the latter generalizing the baseline model for a broader scope for complementarities and addressing prominent shortcomings due to limited mobility bias. We discuss the model results and compare the assumptions given the results that we find. We decompose the log wages across time with the two methods and compare the trends in wage inequality, in particular concerning the increase in Germany between 1995 and 2005. We are able to replicate BLM’s baseline results and find a stark underestimation of sorting of the additive CHK model, likely caused by limited mobility bias. No improvement of fit can be found by allowing for firm-type specific (conditional random) worker effects, i.e. differing complementarities in production. | |||||||
| Lizenz: |  Urheberrechtsschutz | |||||||
| Fachbereich / Einrichtung: | Wirtschaftswissenschaftliche Fakultät » Volkswirtschaftslehre | |||||||
| Dokument erstellt am: | 14.08.2019 | |||||||
| Dateien geändert am: | 14.08.2019 | |||||||
| Promotionsantrag am: | 14.08.2018 | |||||||
| Datum der Promotion: | 18.12.2018 |
