DoCMA Working Papers

Permanent URI for this collection

http://hdl.handle.net/2003/39091

DoCMA ist ein interdisziplinäres Forschungszentrum zur Big-Data-Analyse von Medieninhalten. Ziel ist es, anwendungsorientierte Analyseverfahren für verschiedene Disziplinen zu entwickeln. Im Mittelpunkt steht die Verbindung der sozialwissenschaftlich geprägten Kommunikationswissenschaft als Querschnittsdisziplin mit der mathematisch geprägten Statistik. DoCMA bündelt somit Stärken der TU Dortmund unter neuen Fragestellungen. Bereits seit 2014 kooperieren die beiden beteiligten Professoren. Teams von wissenschaftlichen Mitarbeitern der zwei Fächer arbeiten Hand in Hand. Regelmäßige Workshops, an denen nicht nur Hochschullehrer und Wissenschaftliche Mitarbeiter teilnehmen, sondern gelegentlich auch Master-Studierende, befruchten einen systematischen Austausch.

Browse

Now showing 1 - 13 of 13

The Inflation Attention Cycle
(2023-03) Müller, Henrik; Schmidt, Tobias; Rieger, Jonas; Hornig, Nico; Hufnagel, Lena Marie
An Increasing Sense of Urgency
(2022-08) Müller, Henrik; Rieger, Jonas; Schmidt, Tobias; Hornig, Nico
Vladimir vs. the Virus – a Tale of two Shocks
(2022-05) Müller, Henrik; Rieger, Jonas; Hornig, Nico
Pressure is high – and rising
(2022-05) Müller, Henrik; Rieger, Jonas; Schmidt, Tobias; Hornig, Nico
A German Inflation Narrative
(2022-03) Müller, Henrik; Schmidt, Tobias; Rieger, Jonas; Hufnagel, Lena Maria; Hornig, Nico
In this paper, we present a new indicator to measure the media coverage of inflation. Our Inflation Perception Indicator (IPI) for Germany is based on a corpus of three million articles published by broadsheet newspapers between January 2001 and February 2022. It is designed to detect thematic trends, thereby providing new insights into the dynamics of inflation perception over time. These results may prove particularly valuable at the current juncture, where massive uncertainty prevails due to geopolitical conflicts and the pandemic-related supply-chain jitters. Economists inspired by Shiller (2017; 2020) have called for analyses of economic narratives to complement econometric analyses. The IPI operationalizes such an approach by isolating inflation narratives circulating in the media. Methodically, the IPI makes use of RollingLDA (Rieger et al. 2021), a dynamic topic modeling approach refining the rather static original LDA (Blei et al. 2003) to allow for changes in the model’s structure over time. By modeling the process of collective memory, where experiences of the past are partly overwritten and altered by new ones and partly sink into oblivion, RollingLDA is a potent tool to capture the evolution of economic narratives as social phenomena. In addition, it is suitable to produce stable time-series, to the effect that the IPI can be updated frequently. Our initial results show a narrative landscape in turmoil. Never in the past two decades has there been such a broad shift in inflation perception, and therefore, possibly, in inflation expectations. Also, second-round effects, such as significant wage demands, that have not played a major role in Germany for a long time, seem to be in the making. Towards the end of the time horizon, raw material prices are high on the agenda, too, triggered by the Russian war against Ukraine and the ensuing sanctions against the aggressor. We would like to encourage researchers to use our data and are happy to share it on request.
Text mining methods for measuring the coherence of party manifestos for the German federal elections from 1990 to 2021
(2021-09-14) Jentsch, Carsten; Mammen, Enno; Müller, Henrik; Rieger, Jonas; Schötz, Christof
"Riders on the Storm"
(2021-07) Müller, Henrik; Rieger, Jonas; Hornig, Nico
In this paper we update our Uncertainty Perception Indicator (UPI) with data from the first quarter of 2021. UPI values have declined in recent quarters. At first glance this might come as a surprise, with the third Corona wave having less of an impact than the first one, even though the former was several magnitudes bigger than the latter in Germany. This result underscores the difference between perception and actual impact: a shock hits the hardest when it first occurs because by then its nature is still unknown. As shown in previous versions of the UPI, uncertainty has mainly been fed by the political sphere since the 2010s. Towards the end of our observation period, however, uncertainty from the international and European political spheres is declining, while German domestic politics is on the rise. The end of Angela Merkel's chancellorship marks the end of a long period of relative political stability. Without her in the race the outcome of German federal elections in September is hardly predictable. Whatever coalition may succeed, it is likely that any future government will engineer a shift in (economic) policy. The potential strength of this "election uncertainty effect" is evident in our data. An update of our Fear Gauge shows profound shift in public discourse in Germany. With the pandemic in retreat for now, climate change and the question to what extent policies should follow science (whether on pandemics or global warming) are taking center stage in Germany. Looking ahead, we expect UPI values to rise again as the federal elections loom and the economic and political consequences of the pandemic (e. g. higher debt levels) become apparent. Uncertainty shocks tend to come in waves. Given the severity of the Corona pandemic, a host of difficulties – ranging from unexpected inflation to debt crises to geostrategic tensions – are possibly in the making.
“We’re rolling”
(2021-03) Müller, Henrik; Rieger, Jonas; Hornig, Nico
In this paper, we present a new dynamic topic modeling method to build stable models and consistent time series. We call this new method RollingLDA. It has the potential to overcome several difficulties researchers, who use unsupervised probabilistic topic models, have grappled with: namely the problem of arbitrary selection, which is aggravated when models are to be updated with new sequences of data. RollingLDA is derived by combining the LDAPrototype approach (Rieger, Jentsch and Rahnenführer, 2020) with an implementation that uses preceding LDA results as an initialization for subsequent quarters, while allowing topics to change over time. Squaring dual-process theory, employed in Behavioral Economics (Kahneman, 2011), with the evolving theory of Economic Narratives (Shiller, 2017), RollingLDA is applied to the measurement of economic uncertainty. The new version of our Uncertainty Perception Indicator (UPI), based on a newspaper corpus of 2.8 million German newspaper articles, published between 1 January 2001 and 31 December 2020, proves indeed capable of detecting an uncertainty narrative. The narrative, derived from the thorough quantitative-qualitative analysis of a key-topic of our model, can be interpreted as collective memory of past uncertainty shocks, their causes and the societal reactions to them. The uncertainty narrative can be seen as a collective intangible cultural asset (Haskel and Westlake, 2017), accumulated in the past, informing the present and potentially the future, as the story is being updated and partly overwritten by new experiences. This concept opens up a fascinating new field for future research. We would like to encourage researchers to use our data and are happy to share it on request.
Economic Policy Uncertainty Index
(2021-02) Brandt, Richard
corona100d
(2021-02) Rieger, Jonas; Nordheim, Gerret von
“For the times they are a-changin’”
(2021-02) Müller, Henrik; Hornig, Nico; Rieger, Jonas
This paper deals with the problem of deriving consistent time-series from newspaper contentbased topic models. In the first part, we recapitulate a few our own failed attempts, in the second one, we show some results using a twin strategy, that we call prototyping and seeding. Given the popularity news-based indicators have assumed in econometric analyses in recent years, this seems to be a valuable exercise for researchers working on related issues. Building on earlier writings, where we use the topic modelling approach Latent Dirichlet Allocation (LDA) to gauge economic uncertainty perception, we show the difficulties that arise when a number of one-shot LDAs, performed at different points in time, are used to produce something akin of a time-series. The models’ topic structures differ considerably from computation to computation. Neither parameter variations nor the accumulation of several topics to broader categories of related content are able solve the problem of incompatibleness. It is not just the content that is added at each observation point, but the very properties of LDA itself: since it uses random initializations and conditional reassignments within the iterative process, fundamentally different models can emerge when the algorithm is executed several times, even if the data and the parameter settings are identical. To tame LDA’s randomness, we apply a newish “prototyping” approach to the corpus, upon which our Uncertainty Perception Indicator (UPI) is built. Still, the outcomes vary considerably over time. To get closer to our goal, we drop the notion that LDA models should be allowed to take various forms freely at each run. Instead, the topic structure is fixated, using a “seeding” technique that distributes incoming new data to our model’s existing topic structure. This approach seems to work quite well, as our consistent and plausible results show, but it is bound to run into difficulties over time either.
"I heard the News today, oh Boy"
(2020) Müller, Henrik; Hornig, Nico
News-based indicators are in vogue in economics. But they tend to be applied with little consideration for the properties of news itself. In this paper, we try to shed light on the nature of this type of data. Drawing from established findings in communication science and journalism studies we argue that news-based indicators should be taken with a pinch of salt, since news is a somewhat biased representation of political and social reality. Contrary to economics and other social sciences, journalism tends to be driven by outliers, the outrageous, and the outraged. This structural dissonance between journalism and other disciplines needs to be born in mind when dealing with news content as data, and it is of particular concern in the context of economic developments. While economics and statistics are inherently backward looking, trying to make sense of the (immediate) past using models and probability distributions derived from bygone observations, journalism is about the present, and sometimes about the future. What’s going on right now? And where does it lead us? Seeking answers to these questions makes news a valuable data input, as a measure of what drives society at a given point in time. We show how taking the properties of news into consideration influences the entire process of large-scale news analysis. As an example, we update our Uncertainty Perception Indicator (Müller and Hornig 2020), setting it on a firmer footing by enlarging the newspaper corpus considerably. The new version of the UPI for Germany yields some remarkable results. At the trough of the Covid-19-induced economic crisis in Q2 of 2020, the overall indicator already decreased considerably, although it stayed at elevated levels. Deconstructing the UPI by applying the topic modelling approach Latent Dirichlet Allocation (LDA), shows that the coverage of the pandemic has merged with the issue of climate change and its mitigation. In the past decade or so incalculable politics was the main driver of economic uncertainty perception. Now truly exogenous developments, neither elicited by the economy nor by politics, come to the fore, adding to the sense of an inherently unstable world.
Expecting the Unexpected
(2020-06-17) Müller, Henrik; Hornig, Nico
The phenomenon of economic uncertainty has attracted considerable attention in recent years. New indicators have been introduced aiming at measuring uncertainty and its potential economic consequences. Still, the Corona pandemic has hit the world economy virtually out of the blue. In this paper, we argue that, while it is clear that true uncertainty, by definition, cannot be forecasted, better early warning systems could be built. To further this goal, we propose a new taxonomy of economic uncertainty and construct a news-based indicator that captures different kinds of uncertainty, some of which may precede others. If we are able to detect the preludes of an uncertainty shock, we may be able to gauge its size and potential economic impact early on. In earlier writings (Müller et al. 2018, Müller 2020a) we demonstrated the feasibility of Latent Dirichlet Allocation (LDA) for gauging uncertainty. Here, we base our analysis on an enhanced data set, a broader query, and we propose a routine to scan the recent past for new sources of uncertainty. Based on a text corpus of more than 750.000 newspaper articles published since 2008, we run several topic models of the LDA type. As an unsupervised text mining technique LDA has the potential to make economic indicators more sensitive to hitherto unknown – or overlooked – factors of economically relevant uncertainty. Our results are preliminary, yet encouraging. The notion that economic uncertainty comes in three types, two of which, market-based and economic policy uncertainty, may reinforce one another, while the third type is truly exogeneous, is broadly supported by our empirical approach. The Uncertainty Perception Indicator (UPI) is able to shed light on the links between the three categories of uncertainty and is systematically open to new developments; it is designed to detect not merely known unknowns (e.g. fiscal and monetary policy, trade policy, regulation), but also surprising unknowns (e.g. technological, ecological, social changes).

Browse

Recent Submissions