This dissertation presents three essays that focus on various aspects pertaining to the management of online platforms, defined as "digital services that facilitate interactions between two or more distinct, but interdependent sets of users (whether firms or individuals) who interact through the service via the Internet" (OECD, 2019). The interactions benefit both the users and the platform. Managing online platforms involves developing strategies for one or more of three value adding functions: (a) lowering search costs for the parties connecting through the platform, (b) providing a technology infrastructure that facilitates transactions at scale by sharing both demand and supply side costs; and (c) locating other audiences or consumers for the output that results from the transaction. The platform manager must manage these value adding functions. Thus, one important management task is to recognize potential asymmetries in the economic and/or psychological motivations of the transacting parties connected through the platform. In this dissertation, I empirically examine these issues in greater detail. The first essay, "Incentivizing User-Generated Content—A Double-Edged Sword: Evidence from Field Data and a Controlled Experiment," addresses the conundrum faced by online platform managers interested in crowdsourcing user-generated content (UGC) in prosocial contexts. The dilemma stems from the fact that offering monetary incentives to stimulate UGC contributions also has a damping effect on peer approval, which is an important source of non-monetary recognition valued by UGC contributors in prosocial contexts. The second essay, "Matching and Making in Matchmaking Platforms: A Structural Analysis," examines matchmaking platforms, focusing specifically on the problem of misaligned incentives between the platform and the agents. Based on data from the Ultimate Fighting Championship (UFC) on fighter characteristics, and pay-per-view revenues associated with specific bouts, we identify the potential for conflicts of interest and examine strategies that may be used to mitigate such problems. The third essay, "Matching and Making in Matching Markets: A Managerial Decision Calculus," extends the empirical model and analytical work to a class of commonly encountered one-sided matching market problems. It provides the conceptual outline of a decision calculus that allows managers to explore the revenue and profitability implications of adaptive changes to the tier structures and matching algorithms. / Doctor of Philosophy / The 21st century has witnessed the rise of the platform economy. Consumers routinely interact with online platforms ways in their day to day activities. For instance, they interact with platforms such as Quora, StackOverflow, Uber, and Airbnb to name only a few. Such platforms address a variety of needs starting from providing users with answers to a variety of questions to matching them with a range of service providers (e.g., for travel and dining needs). However, the rapid growth of the platform economy has created a knowledge gap for both consumers and platforms. The three essays in this dissertation attempt to contribute to the literature in this area. The first essay, "Incentivizing User-Generated Content—A Double-Edged Sword: Evidence from Field Data and a Controlled Experiment," examines how crowdsourcing contests influence the quantity and quality of user-generated content (UGC). Analyzing data from the popular question and answer website Quora, we find that offering monetary incentives to stimulate UGC contributions increases contributions but also has a simultaneous damping effect on peer endorsement, which is an important source of non-monetary recognition for UGC contributors in prosocial contexts. The second essay, "Matching and Making in Matchmaking Platforms: A Structural Analysis," examines matchmaking platforms, focusing on the problem of misaligned incentives between the platform and the agents. Based on data from the Ultimate Fighting Championship (UFC) on fighter characteristics, and pay-per-view revenues associated with specific bouts, we identify the potential for conflicts of interest and examine strategies that may be used to mitigate such problems. The third essay, "Matching and Making in Matching Markets: A Managerial Decision Calculus," extends the empirical model and analytical work to a class of commonly encountered one-sided matching market problems. It provides the conceptual outline of a decision calculus that allows managers to explore the revenue and profitability implications of adaptive changes to the tier structures and matching algorithms.

Diese Dissertation umfasst drei empirische Analysen regionaler Arbeitsmärkte in Deutschland. Wir wenden dabei Methoden der räumlichen Ökonometrie auf regionale Arbeitsmarktdaten an, um der räumlichen Struktur von Arbeitsmarktaktivitäten Rechnung zu tragen. In der ersten Analyse schlagen wir ein räumliches Paneldatenmodell zur Untersuchung deutscher Matchingfunktionen vor. Mit Hilfe dieses Modells sollen verzerrte und ineffiziente Koeffizientenschätzungen aufgrund räumlicher Abhängigkeiten vermieden werden. Wir zeigen, dass das Vernachlässigen der räumlichen Struktur zu nach oben verzerrten Matchingkoeffizienten führt. Das Ziel der zweiten Untersuchung ist es, Bestimmungsfaktoren für regionale Unterschiede in Arbeitslosigkeit zu identifizieren. Unsere Ergebnisse zeigen, dass ein räumlich und zeitlich dynamisches Paneldatenmodell am besten für diese Fragestellung geeignet ist. Zudem zeigen unsere Ergebnisse, dass die regionalen Unterschiede in der deutschen Arbeitslosigkeit einen Ungleichgewichtszustand darstellen. Diese Erkenntnis kann als Argument für politische Interventionen dienen. In der letzten Analyse wenden wir uns der räumlichen Gewichtungsmatrix zu, der eine zentrale Bedeutung in räumlichen Modellen zukommt. Auf Basis einer empirischen Analyse wollen wir die Definition von räumlichen Gewichten untersuchen und ermitteln Faktoren, die die räumlichen Abhängigkeiten auf Arbeitsmärkten bestimmen. Dabei untersuchen wir sowohl unterschiedliche Dimensionen ökonomischer als auch geographische Distanzen als Wirkungskanal räumlicher Abhängigkeit. Für die Entscheidung, welche dieser Distanzdimensionen einen Einfluss auf die räumlichen Relationen hat, verwenden wir ein räumlich-autoregressives Modell höherer Ordnung. Unsere Ergebnisse zeigen, dass geographische Distanz alleine nicht ausreicht, um die räumlichen Interdependenzen zwischen regionalen Arbeitsmärkten zu erklären, sondern auch Dimensionen ökonomischer Distanz einen signifikanten Erklärgehalt haben. / In this dissertation, we present different empirical analyses of regional labor markets in Germany. To account for the spatial structure of labor market activities, we apply spatial econometric methods to regional labor market data. In the first analysis, we propose a spatial panel model for German matching functions to avoid possibly biased and inefficient estimates due to spatial dependence. Based on an official data set, we show that neglecting spatial dependencies in the data results in upward-biased coefficients. Furthermore, our results suggest that a dynamic modeling is more appropriate for matching functions than a static approach. In the second analysis, we study determinants for regional differences in unemployment rates. We specify a spatial panel model to avoid biased and inefficient estimates due to spatial dependence. The study covers the whole of Germany as well as East and West Germany separately. Our results suggest that a spatial dynamic panel model is the best model for this analysis. Moreover, we find that German regional unemployment is of disequilibrium nature, which justifies political interventions. Finally, we study the spatial weights matrix which is the key component in spatial econometric models. We investigate empirically the issue of defining spatial weights in labor market applications and propose factors driving spatial dependence in regional labor markets. In addition to geographic distance, we consider different dimensions of economic distance as transmission channel of spatial dependence. To decide which factors influence spatial dependence in labor markets, we apply a higher-order spatial autoregressive model to data on regional labor markets in Germany. Our results suggest that geographic distance does not capture the spatial dependence between regional labor markets sufficiently but economic distance needs to be considered as well.