Der Wert der Worte – Wie digitale Methoden helfen, Kommunikations- und Wirtschaftswissenschaft zu verknüpfen

The value of words—how digital methods help to connect communication science and economics

Zusammenfassung

Die Anwendung digitaler Methoden bietet die Chance, Kommunikationswissenschaften und Ökonomik enger zu verknüpfen. So besteht in der Konjunktur- und der Finanzmarktforschung seit einigen Jahren ein wachsendes Interesse an der Einbeziehung von Medieninhalten, allerdings meist ohne Rückgriff auf die Kommunikationswissenschaften. Dieser Beitrag gibt einen Überblick über Ansätze zur Einbeziehung von Medieninhalten in der Wirtschaftsforschung und stellt einen neuen, multidimensionalen Indikator vor, der wirtschaftspolitische Unsicherheit auf der Grundlage des Topic-Modelling-Verfahrens Latent Dirichlet Allocation (LDA) misst: den Uncertainty Perception Indicator (UPI). Auf dieser Basis zeigen wir, wie sich der in der Ökonomik populäre Begriff des „Narrativs“ operationalisieren und mit dem kommunikationswissenschaftlichen Konzept des Frames verknüpfen lässt.

Abstract

The use of digital methods offers a chance to connect communication science with economics. In recent years, a growing body of research in economics has turned its attention to media content, assuming that journalistic coverage contains hitherto neglected information relevant for business cycles or financial market movements. Interestingly, these approaches largely ignore communication science’s established theories and empirical findings. This paper aims at building a bridge between the two disciplines. Its contribution is threefold: a) it provides an overview of the most important approaches in economics that incorporate media content; b) it operationalizes the concept of the “narrative”, as it is used in economics, and distinguishes it from the concept of the “frame”, essential in communication science; c) exemplifying our approach, we present a new Uncertainty Perception Indicator (UPI) based on the topic modeling method Latent Dirichlet Allocation (LDA), that enables us to isolate different factors of economic policy uncertainty contained in media coverage.

Economic studies treat journalistic media content as a proxy for sentiment prevalent in society. Typically, they rely on frequency analyses of certain keywords, like “recession” or “inflation”. Even more sophisticated approaches, such as Shiller (2017), who calls for establishing a new branch of “narrative economics”, or Baker et al. (2016), who construct a comprehensive set of media-based indicators, make no or little reference to communication science. This neglect could be discounted as pure ignorance, but this misses the point. Being a predominantly empirical discipline today, economics relies on long time-series of data, that have not been available for media content, a gap rendering the two disciplines largely incompatible.

The gap is also reflected in terminology. “Frame” is a major analytical concept in communication science, while the term “narrative” has become in vogue in economics. Although both concepts are closely connected, they are rarely properly distinguished from each other. “Frame” can be considered as a rather static concept that applies during a limited period of time. “Narrative”, in contrast, implies dynamic properties, i. e., the sorting of events, causes and effects over time, that explain how the current state of the world has come about, as stressed by Tenenboim-Weinblatt et al. (2016).

In this paper, we propose a synergetic concept. Following Entman (1993), a media frame contains four elements: a) a problem definition, b) a problem diagnosis, c) a moral judgement, and d) possible remedies. We augment this approach by adding two more elements. According to our definition, a media narrative comprises a frame, or several ones, plus e) one or several protagonists—persons, institutions, or social groupings (nations, classes, etc.)—, whose relationships are (often) antagonistic and may change over time; and f) events, that are chronologically integrated and that are (often) assumed to constitute causal relationships. To put it metaphorically: a frame is to a narrative what a still photo is to a movie. Both are valuable concepts; the still photo shows more details, while the movie provides a contextualization over time.

Topic models like LDA are valuable tools for the measurement of media narratives. The probabilistic approach enables researchers to conduct what may be called “macro-content analyses”, an exercise that focuses on average reporting patterns in large text corpora and can be translated into numerical time-series, thereby facilitating compatibility with empirical economics. Based on a topic’s frequency analysis, its top words and top articles, “mean media narratives” can be formulated, that integrate certain events, protagonists and frames.

In our case study, we exemplify this concept by applying it to an indicator that is currently popular in economics, the Economic Policy Uncertainty Index (Baker et al. 2016). The EPU aims at capturing political developments that are exogenous to economic models and therefore unpredictable. Essentially, the indicator is based on the counts of articles containing a set of search words, such as “uncertain”, “economic” as well as institutions like the European Central Bank. Using identical search words as the EPU for Germany, we construct a similar corpus for the years 1994 to 2017. By conducting an LDA-based analysis, we are able to extract additional relevant information from the data. In particular, the evolution of different uncertainty factors and their development over time can be detected.

Our Uncertainty Perception Indicator (UPI) contains six relevant news topics that are highly relevant for market developments: central banks, the national government, international politics, the business cycle, companies, and society. While the EPU merely shows how often uncertainty concerning economic policy is mentioned in the media, the UPI also indicates the origins of uncertainty. By grouping the six topics into three analytical categories—governments, markets, and society—we find a distinct break in the time-series. Before the financial crisis of 2008, the perception of uncertainty was rather balanced between the three factors. Since then, however, economic uncertainty has mainly been driven by political actors, most prominently by central banks. The corresponding narratives are a two-chapter story: in the first part, up to 2008, stable financial markets and smoothed business cycles prevailed, making central banking a rather straight-forward task. The second part is characterized by multiple crises, leaving central banks as dominant actors, that intervened with unconventional measures. Thereby, they became stabilizing forces, but at the same time sources of uncertainty with respect to the timing and the impact of these measures.

This is a preview of subscription content, log in to check access.

Abb. 1
Abb. 2
Abb. 3
Abb. 4
Abb. 5
Abb. 6
Abb. 7
Abb. 8

Notes

  1. 1.

    Verschiedene Arbeiten beschreiben Journalisten aus der Perspektive der Cultural Studies als Storyteller eines gesellschaftlichen Narrativs (vgl. Lünenborg 2005; Roeh 1989; Tuchman 1976), das Identität stiftet, indem es bestimmte Weltsichten und Werte vermittelt (vgl. Bird und Dardenne 2009; Hickethier 1997).

  2. 2.

    Entsprechend konzentrieren sich klassische Inhaltsanalysen auf Zeitpunkte oder kurze Zeitabschnitte, an denen die Berichterstattung erhoben und verglichen wird. Dieser Ansatz findet seine Entsprechung in der Ökonomik in der komparativ-statischen Analyse, die verschiedene Gleichgewichtszustände miteinander vergleicht, jedoch die Anpassungswege zwischen zwei Gleichgewichten nicht zu beschreiben in der Lage ist. Betrachtet man das Narrativ als temporalen Begriff, wird umso deutlicher, warum die Zeitreihen-fokussierte Ökonomik derartige Affinität für den Begriff entwickelt hat.

  3. 3.

    So findet sich in der Gemeinschaftsdiagnose der deutschen Wirtschaftsforschungsinstitute vom April 2018 der Begriff Unsicherheit 15-mal, jeweils im Kontext von politischen Prozessen. Insbesondere: die künftige Ausrichtung der US-Wirtschaftspolitik, zumal der Handelspolitik; die Umsetzung der im deutschen Koalitionsvertrag angekündigten zusätzlichen Ausgaben und Entlastungen; die Auswirkungen dieser Unsicherheit auf die Finanzmärkte und auf die Stimmung bei den Unternehmen; Regulierungen des französischen Arbeitsmarkts (vgl. Projektgruppe Gemeinschaftsdiagnose 2018).

  4. 4.

    Für eine detaillierte Beschreibung des EPU siehe www.policyuncertainty.com.

  5. 5.

    Bei einer LDA, die einen langen Zeitraum (hier: 24 Jahre) umspannt, ist es möglich, dass die relative Größe von Topics sich im Zeitablauf verändert, weil das dahinterliegende Modell Veränderungen in der Berichterstattung nicht hinreichend erfasst. Der Rückgang der Größe des Topics „Regierung“ könnte somit modellimmanent begründet sein. Dagegen spricht, dass die Rangfolge der Top-Words mit der Rangfolge der Regierungszeiten übereinstimmt. So waren die Unionsparteien (Top-Words „cdu“, „csu“, „union“) im betrachteten Zeitraum in insgesamt 17 Jahren an Bundesregierungen beteiligt, die SPD in 15 Jahren, die FDP in neun Jahren, die Grünen in sieben Jahren. Einzig die SPD ist, so gesehen, in der Top-Word-Liste überrepräsentiert (was mit der starken Stellung der SPD im Bundesrat in der Endphase der Ära Kohl erklärbar sein dürfte). Auch die Dauer der Kanzlerschaften findet in der Top-Word-Rangfolge ihre Entsprechung: Zwischen 1994 und 2017 war Merkel zwölf Jahre im Amt, Schröder sieben, Kohl fünf Jahre.

  6. 6.

    Mit „Taper Tantrum“ wurde die Unruhe an den Finanzmärkten, zumal in Schwellenländern, bezeichnet, die entstand, nachdem die Fed angekündigt hatte, ihr Anleihekaufprogramm zurückzufahren.

  7. 7.

    Die Ankündigung wurde später von der EZB im „Outright Monetary Transactions“-Programm (OMT) formalisiert.

  8. 8.

    Das ist insofern analyseadäquat, als sich Sanktionen gegen Russland nur marginal auf die wirtschaftliche Gesamtentwicklung in Deutschland auswirken.

Literatur

  1. Akerlof, G. A., & Shiller, R. J. (2009). Animal spirits. How human psychology drives the economy, and why it matters for global capitalism. Princeton: Princeton University Press.

    Google Scholar 

  2. Akerlof, G. A., & Shiller, R. J. (2015). Phishing for fools. The economics of manipulation and deception. Princeton: Princeton University Press.

    Google Scholar 

  3. Akerlof, G. A., & Snower, D. J. (2016). Bread and bullets. Kiel working paper no. 2022.

    Google Scholar 

  4. Ammann, M., Frey, R., & Verhofen, M. (2012). Do newspaper articles predict aggregate stock returns? Working papers on finance no. 2012/4. St. Gallen: Swiss Institute of Banking and Finance.

    Google Scholar 

  5. Baker, S. R., Bloom, N., & Davis, S. J. (2016). Measuring economic policy uncertainty. The Quarterly Journal of Economics, 131, 1593–1636.

    Article  Google Scholar 

  6. Bird, E., & Dardenne, R. W. (2009). Rethinking news and myth as storytelling. In T. Hanitzsch & K. Wahl-Jorgensen (Hrsg.), The handbook of journalism studies (S. 205–217). New York: Routledge.

    Google Scholar 

  7. Blei, D. M., Ng, A. Y., & Jordan, M. I. (2003). Latent Dirichlet allocation. Journal of Machine Learning Research, 3, 993–1022. http://www.jmlr.org/papers/volume3/blei03a/blei03a.pdf.

    Google Scholar 

  8. Boumans, J. W., & Trilling, D. (2016). Taking stock of the toolkit: an overview of relevant autmated content analysis approaches and techniques for digital journalism scholars. Digital Journalism, 4(1), 8–23.

    Article  Google Scholar 

  9. Brosius, H.-B., & Weimann, G. (1996). Who sets the agenda: agenda-setting as a two-step flow. Communication Research, 23, 561–580.

    Article  Google Scholar 

  10. Brunnermeier, M. K., James, H., & Landau, J.-P. (2016). The Euro and the battle of ideas. Princeton: Princeton University Press.

    Google Scholar 

  11. Burscher, B., Odijk, D., Vliegenthart, R., de Rijke, M., & de Vreese, C. H. (2014). Teaching the computer to code frames in news: comparing two supervised machine learning approaches to frame analysis. Communication Methods and Measures, 8(3), 190–206.

    Article  Google Scholar 

  12. Bybee, J. L., & Hopper, P. J. (2001). Introduction to frequency and the emergence of linguistic structure. In J. L. Bybee & P. J. Hopper (Hrsg.), Frequency and the emergence of linguistic structure. Amsterdam: John Benjamins.

    Google Scholar 

  13. Calomiris, C. W., & Mamaysky, H. (2018). How news and its context drive risk and returns around the world. NBER working paper 24430.

    Google Scholar 

  14. Carroll, C. D. (2003). Macroeconomic expectations of households and professional forecasters. The Quarterly Journal of Economics, 118(1), 269–298.

    Article  Google Scholar 

  15. Cohen, B. (1963). The press and foreign policy. New York: Harcourt.

    Google Scholar 

  16. Czarniawska, B. (2004). Narratives in social science research. London: SAGE.

    Google Scholar 

  17. De Boef, S., & Kellstedt, P. M. (2004). The political (and economic) origins of consumer confidence. American Journal of Political Science, 48, 633–649.

    Article  Google Scholar 

  18. De Vreese, C. H. (2010). Framing the economy: effects of journalistic news frames. In P. D’Angelo & J. Kuypers (Hrsg.), Doing framing analysis: empirical and theoretical perspectives (S. 187–214). New York: Routledge.

    Google Scholar 

  19. DiMaggio, P., Nag, M., & Blei, D. M. (2013). Exploiting affinities between topic modeling and the sociological perspective on culture: application to newspaper coverage of U.S. government arts funding. Poetics, 41, 570–606.

    Article  Google Scholar 

  20. Doms, M., & Morin, N. (2004). Consumer sentiment, the economy, and the news media. Finance and economics discussion series 2004-51. Washington: Divisions of Research & Statistics and Monetary Affairs, Federal Reserve Board.

    Google Scholar 

  21. Entman, R. M. (1993). Framing: toward clarification of a fractured paradigm. Journal of Communication, 43(4), 51–58.

    Article  Google Scholar 

  22. European Economic Advisory Group (EEAG) (2018). EEAG report on the European economy. München: CESifo.

    Google Scholar 

  23. Fogarty, B. J. (2005). Determining economic news coverage. International Journal of Public Opinion Research, 17, 149–172.

    Article  Google Scholar 

  24. Gamson, W. A., Croteau, D., Hoynes, W., & Sasson, T. (1992). Media images and the social construction of reality. Annual Review of Sociology, 18, 373–393.

    Article  Google Scholar 

  25. Ghanem, S. I. (1997). Filling in the tapestry: the second level of agenda setting. In M. E. McCombs, D. L. Shaw & D. H. Weaver (Hrsg.), Communication and democracy (S. 3–14). Mahwah: Lawrence Erlbaum.

    Google Scholar 

  26. Grimmer, J., & Stewart, B. M. (2013). Text as data: the promise and pitfalls of automatic content analysis methods for political texts. Political Analysis, 21, 267–297.

    Article  Google Scholar 

  27. Günther, E., & Quandt, T. (2016). Word counts and topic models. Digital Journalism, 4(1), 75–88.

    Article  Google Scholar 

  28. Györffy, D. (2013). Institutional trust and economic policy. Lessons from the history of the Euro. Budapest: Central European University Press.

    Google Scholar 

  29. Haller, H. B., & Norpoth, H. (1997). Reality bites: news exposure and economic opinion. Public Opinion Quarterly, 61, 555–576.

    Article  Google Scholar 

  30. Harris, Z. S. (1951). Methods in structural linguistics. Chicago: University of Chicago Press.

    Google Scholar 

  31. Hepp, A. (2016). Kommunikations- und Medienwissenschaft in datengetriebenen Zeiten. Publizistik, 61, 225–246.

    Article  Google Scholar 

  32. Hickethier, K. (1997). Das Erzählen der Welt in den Fernsehnachrichten. Überlegungen zu einer Narrationstheorie der Nachricht. Rundfunk und Fernsehen, 45, 5–18.

    Google Scholar 

  33. Iyengar, S., & Kinder, D. R. (1987). News that matters: television and American opinion. American politics and political economy series. Chicago: University of Chicago Press.

    Google Scholar 

  34. Keynes, J. M. (1936). The general theory of employment, interest and money. London: Palgrave Macmillan.

    Google Scholar 

  35. Kholodilin, K., Thomas, T., & Ulbricht, D. (2014). Do media data help to predict German industrial production? Dice discussion paper no 149.

    Google Scholar 

  36. Kholodilin, K., Kolmer, C., Thomas, T., & Ulbrich, D. (2015). Asymmetric perceptions of the economy: media, firms, consumers, and experts. DIW discussion paper 1490.

    Google Scholar 

  37. Knight, F. (1921). Risk, uncertainty and profit. Cambridge: The Riverside Press.

    Google Scholar 

  38. Koppers, L., Rieger, J., Boczek, K., & von Nordheim, G. (2018). tosca (tools for statistical content analysis). https://cran.r-project.org/web/packages/tosca/index.html Zugegriffen: 10. Oktober 2018.

    Google Scholar 

  39. Larsen, V. H., & Thorsrud, L. A. (2015). The value of news. CAMP working paper series. No 6/2015.

    Google Scholar 

  40. Linden, F. (1982). The consumer as forecaster. Public Opinion Quarterly, 46, 353–360.

    Article  Google Scholar 

  41. Lischka, J. A. (2016). Economic news, sentiment, and behavior. Wiesbaden: Springer.

    Google Scholar 

  42. Lünenborg, M. (2005). Journalismus als kultureller Prozess. Zur Bedeutung von Journalismus in der Mediengesellschaft. Ein Entwurf. Wiesbaden: Springer.

    Google Scholar 

  43. Maier, D., Waldherr, A., Miltner, P., Wiedemann, G., Niekler, A., Keinert, A., et al. (2018). Applying LDA topic modeling in communication research. Toward a valid and reliable methodology. Communication Methods and Measures, 54(10), 1–26.

    Google Scholar 

  44. Matthes, J. (2006). The need for orientation towards news media: revising and validating a classic concept. International Journal of Public Opinion Research, 18, 422–444.

    Article  Google Scholar 

  45. McComas, K., & Shanahan, J. (1999). Telling stories about global climate change. Measuring the impact of narratives on issue cycles. Communuication Research, 26, 30–57.

    Article  Google Scholar 

  46. McCombs, M. E., & Shaw, D. L. (1972). The agenda-setting function of mass media. Public Opinion Quarterly, 36, 176–187.

    Article  Google Scholar 

  47. Müller, H., Porcaro, G., & von Nordheim, G. (2018). Tales from a crisis: diverging narratives of the euro area. Bruegel policy contribution, issue no. 03.

    Google Scholar 

  48. Nadeau, R., Niemi, R. G., Fan, D. P., & Amato, T. (1999). Elite economic forecasts, economic news, mass economic judgments, and presidential approval. Journal of Politics, 61, 109–135.

    Article  Google Scholar 

  49. Neuman, R. W., Guggenheim, L., Jang, M. S., & Bae, S. Y. (2014). The dynamics of public attention: agenda-setting theory meets big data. Journal of Communication, 64, 193–214.

    Article  Google Scholar 

  50. Noelle-Neumann, E., & Mathes, R. (1987). The ‘event as event’ and the ‘event as news’: the significance of ‘consonance’ for media effects research. European Journal of Communication, 2, 391–414.

    Article  Google Scholar 

  51. Projektgruppe Gemeinschaftsdiagnose (2018). Deutsche Wirtschaft im Boom – Luft wird dünner. Gemeinschaftsdiagnose 1/2018. http://gemeinschaftsdiagnose.de/wp-content/uploads/2018/04/GD_F18_Langfassung_unkorrigiert.pdf. Zugegriffen: 20. Apr. 2018.

    Google Scholar 

  52. Puschmann, C., & Scheffler, T. (2016). Topic modeling for media and communication research. A short primer. HIIG discussion paper series 5. (S. 1–17). https://papers.ssrn.com/sol3/Delivery.cfm/SSRN_ID2837801_code1779785.pdf?abstractid=2836478&mirid=1

    Google Scholar 

  53. Richter, R. (2015). Essays in institutional economics. Heidelberg: Springer.

    Google Scholar 

  54. Roeh, I. (1989). Journalism as storytelling, coverage as narrative. American Behavioral Scientist, 33(2), 162–168.

    Article  Google Scholar 

  55. Shiller, R. J. (2000). Irrational exuberance. Princeton: Princeton University Press.

    Google Scholar 

  56. Shiller, R. J. (2017). Narrative economics. American Economic Review, 107(4), 967–1004.

    Article  Google Scholar 

  57. Soroka, S. N. (2006). Good news and bad news: asymmetric responses to economic information. Journal of Politics, 68, 372–385.

    Article  Google Scholar 

  58. Strippel, C., Bock, A., Katzenbach, C., Mahrt, M., Merten, L., Nuernbergk, C., Pentzold, C., Puschmann, C., & Waldherr, A. (2018). Die Zukunft der Kommunikationswissenschaft ist schon da, sie ist nur ungleich verteilt. Eine Kollektivreplik. Publizistik, 63, 11–27.

    Article  Google Scholar 

  59. Tenenboim-Weinblatt, K., & Neiger, M. (2015). Print is future, online is past: cross-media analysis of temporal orientations in the news. Communication Research, 42, 1047–1067.

    Article  Google Scholar 

  60. Tenenboim-Weinblatt, K., Hanitzsch, T., & Nagar, R. (2016). Beyond peace journalism: reclassifying conflict narratives in the Israeli news media. Journal of Peace Research, 53(2), 151–165.

    Article  Google Scholar 

  61. Tetlock, P. C. (2007). Giving content to investor sentiment: the role of media in the stock market. The Journal of Finance, 62, 1139–1168.

    Article  Google Scholar 

  62. Thorsrud, L. A. (2016a). Nowcasting using news topics. Big data versus big bank. CAMP working paper series no 6/2016.

    Google Scholar 

  63. Thorsrud, L. A. (2016b). Words are the new numbers: a newsy coincident index of business cycles. CAMP working paper series, no 4/2016.

    Google Scholar 

  64. Tsur, O., Calacci, D., & Lazer, D. (2015). A frame of mind: using statistical models for detection of framing and agenda setting campaigns. In Proceedings of the 53rd annual meeting of the association for computational linguistics and the 7th international joint conference on natural language processing (S. 1629–1638). Stroudsburg: ACL.

    Google Scholar 

  65. Tuchman, G. (1976). Telling stories. Journal of Communication, 26(4), 93–97.

    Article  Google Scholar 

  66. Vosen, S., & Schmidt, T. (2011). Forecasting private consumption: survey-based indicators vs. google trends. Journal of Forecasting, 30, 565–578.

    Article  Google Scholar 

  67. Wanta, W., & Hu, Y.-W. (1994). The effects of credibility, reliance, and exposure on media agenda-setting: a path analysis model. Journalism Quarterly, 71, 90–98.

    Article  Google Scholar 

  68. Wu, H. D., & Coleman, R. (2009). Advancing agenda-setting theory: the comparative strength and new contingent conditions of the two levels of agenda-setting effects. Journalism & Mass Communication Quarterly, 86(4), 775–789.

    Article  Google Scholar 

  69. Wu, H. D., Stevenson, R. L., Hsiao-Chi, C., & Guner, Z. N. (2002). The conditional impact of recession news: a time-series analysis of economic communication in the United States, 1987–1996. International Journal of Public Opinion Research, 14, 19–36.

    Article  Google Scholar 

  70. Wyss, V. (2010). Narration freilegen: Zur Konsequenz der Mehrsystemrelevanz als Leitdifferenz des Qualitätsjournalismus. In K. Imhof, R. Blum & H. Bonfadelli (Hrsg.), Krise der Leuchttürme öffentlicher Kommunikation – Vergangenheit und Zukunft der Qualitätsmedien (S. 31–47). Wiesbaden: Springer.

    Google Scholar 

Download references

Danksagung

Die Autoren danken den Gutachtern und Herausgebern für viele wertvolle Hinweise, die Eingang in dieses Papier gefunden haben.

Author information

Affiliations

Authors

Corresponding author

Correspondence to Prof. Dr. Henrik Müller.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Müller, H., von Nordheim, G., Boczek, K. et al. Der Wert der Worte – Wie digitale Methoden helfen, Kommunikations- und Wirtschaftswissenschaft zu verknüpfen. Publizistik 63, 557–582 (2018). https://doi.org/10.1007/s11616-018-0461-x

Download citation

Schlüsselwörter

  • Topic Modelling
  • Wirtschaftswissenschaften
  • Narrative
  • Framing
  • Unsicherheit

Keywords

  • Topic models
  • Economics
  • Narrative
  • Framing
  • Uncertainty