“Attention and Peer Production of User Generated Content – Evidence from Pseudo-Experiments on Wikipedia”
(2016, in preparation for Submission)
What can additional attention contribute to the high quality and quantity on user generated content networks? Correlations abound in such networks, but it is typically difficult to identify causal effects or merely exogenous sources of variation. This paper proposes to use observable (ex-post) local shocks in attention, such as natural disasters, to analyze how attention contributes to content generation. I analyze the effect of more than 90 pseudo-experimental shocks to attention on content contributions to articles that are linked to shocked articles. I find that 1000 aggregate views generate 1 edit. The target article’s characteristics, such as length, do not influence initial attention, but matter for the conversion of attention to subsequent contributions. However, attention driven contributions seem to add only marginally to already well developed articles, but they do not lead to substantial content additions on short articles.
[Online Appendix – Natural Disasters];
[Online Appendix – Data Description]
“When Private Information Settles the Bill: Money and Privacy in Google ˇApps Market for Smartphone Applications”,
(with Patrick Schulte. 2017, R&R [pdf] [link] )
We shed light on a money-for-privacy trade-off in the market for smartphone applications (”apps”). Developers offer their apps cheaper in return for greater access to personal information, and consumers choose between lower prices and more privacy. We provide evidence for this pattern using data on 300,000 mobile applications which were obtained from the Android Market in 2012 and 2014. We augmented these data with information from Alexa.com and Amazon Mechanical Turk. Our findings show that both the market’s supply and the demand side consider an app’s ability to collect private information, measured by their use of privacy-sensitive permissions: (1) cheaper apps use more privacy-sensitive permissions; (2) installation numbers are lower for apps with sensitive permissions; (3) circumstantial factors, such as the reputation of app developers, mitigate the strength of this relationship. Our results emerge consistently across several robustness checks, including the use of panel data analysis, the use of selected matched ”twin”-pairs of apps and the use of various alternative measures of privacy-sensitiveness.
“Centrality and Content Creation in Networks – The Case of German Wikipedia”
(with Marianne Saam (ZEW), Iassen Halatchliyski (IWM, Tübingen) and George Giorgidze (Uni Tübingen), 2014, published at IEP, 2016)
When contributing content to large and highly structured online platforms like Wikipedia, producers of user-generated content have to decide where to contribute. This decision is expected to depend on the way the content is organized. We analyse whether the hyperlinks on Wikipedia channel the attention of producers towards more central articles. We observe a sample 7, 635 articles belonging to the category economics on the German Wikipedia over 153 weeks and we measure their centrality both within this category and in the network of over one million German Wikipedia articles. Our analysis reveals that an additional link from the observed category is associated with around 140 bytes of additional content and with an increase in the number of authors by 0.5. The relation of links from outside the category to content creation is much weaker. Download
An earlier version was circulated as, “Centrality and Content Creation in Networks – The Case of German Wikipedia”, ZEW Discussion Paper No. 12-053, Mannheim.
“Spillovers in Networks of User Generated Content – Pseudo-Experimental Evidence on Wikipedia”
( [JOB MARKET PAPER], 2014)
Networks are widely believed to generate important spillovers and peer effects. However, quantifying such externalities has traditionally been challenging, because networks often form endogenously. This problem can be circumvented if one can find or create exogenous changes in a network structure. I find such exogeneous variation in the setting of German Wikipedia. Wikipedia prominently advertises a featured article on its main site every day. The added exposure exogeneously increases viewership of this article, while shifts in the viewership of adjacent articles are likely due to their link from the treated article. Through this approach I isolate how the link network causally influences users’ search and contribution behavior. I estimate how attention spills to neighbors through these transient shock in a difference-in-differences analysis. I further develop an extended peer effects model which relaxes the requirement of exogenously given networks to allow estimating the underlying spillover. Advertisements affect neighbors substantially: They increase views on neighbors by almost 70 percent. This, in turn, translates into increased editing activity. My methods apply even if identification through partial overlaps in the network structure fails.An earlier version was circulated as “Spillovers in Networks of User Generated Content – Evidence from 23 Natural Experiments on Wikipedia” and is available as ZEW Disc. Paper 13-098.