|
on Big Data |
| By: | Isabel Hovdahl |
| Abstract: | One of the most important research questions in climate economics is the relationship between temperatures and human mortality. This paper develops a procedure that enables the use of machine learning for estimating the causal temperature-mortality relationship. The machine-learning model is compared to a traditional OLS model, and although both models are capturing the causal temperature-mortality relationship, they deliver very di?erent predictions of the e?ect of climate change on mortality. These di?erences are mainly caused by di?erent abilities regarding extrapolation and estimation of marginal e?ects. The procedure developed in this paper can ?nd applications in other ?elds far beyond climate economics. |
| Keywords: | Climate change, machine learning, mortality |
| Date: | 2019–06 |
| URL: | http://d.repec.org/n?u=RePEc:bny:wpaper:0077&r=all |
| By: | Richard Heuver; Ron TriepelsTriepels |
| Abstract: | Liquidity stress constitutes an ongoing threat to financial stability in the banking sector. A bank that manages its liquidity inadequately might find itself unable to meet its payment obligations. These liquidity issues, in turn, can negatively impact the liquidity position of many other banks due to contagion effects. For this reason, central banks carefully monitor the payment activities of banks in financial market infrastructures and try to detect early-warning signs of liquidity stress. In this paper, we investigate whether this monitoring task can be performed by supervised machine learning. We construct probabilistic classifiers that estimate the probability that a bank faces liquidity stress. The classifiers are trained on a dataset consisting of various payment features of European banks and which spans several known stress events. Our experimental results show that the classifiers detect the periods in which the banks faced liquidity stress reasonably well. |
| Keywords: | Risk Monitoring; Liquidity Stress; Neural Networks; Financial Market Infrastructures; Large-Value Payment Systems |
| JEL: | G32 G33 C45 E42 |
| Date: | 2019–06 |
| URL: | http://d.repec.org/n?u=RePEc:dnb:dnbwpp:642&r=all |
| By: | Johannes von Bloh; Tom Broekel; Burcu Oezgun; Rolf Sternberg |
| Abstract: | Although conventional register and survey data on entrepreneurship have enabled remarkable insights into the phenomenon, the added value has slowed down noticeably over the last decade. There is a need for fresh approaches utilising modern data sources such as Big Data. Until now, it has been quite unknown whether Big Data actually embodies valuable contributions for entrepreneurship research and where it can perform better or worse than conventional approaches. To contribute towards the exploration of Big Data in entrepreneurship research, we use a newly developed dataset based on publications of the German Press Agency (dpa) to explore the relationship between news coverage of entrepreneurship and regional entrepreneurial activity. Furthermore, we apply sentiment analysis to investigate the impact on sentiment of entrepreneurial press releases. Our results show mixed outcomes regarding the relationship between reporting of entrepreneurial events, i.e., media coverage, and entrepreneurial activity in German planning regions. At this stage, our empirical results reject the idea of a strong relationship between actual entrepreneurial activities in regions and the intensity of it being reported. However, the results also imply much potential of Big Data approaches for further research with more sophisticated methodology approaches. Our paper provides an entry point into Big Data usage in entrepreneurship research and we suggest a number of relevant research opportunities based on our results. |
| Keywords: | entrepreneurship, media coverage, mass media, Big Data, sentiment analysis, GEM, entrepreneurial ecosystem, region, news data |
| JEL: | C8 L26 R12 |
| Date: | 2019–06 |
| URL: | http://d.repec.org/n?u=RePEc:egu:wpaper:1920&r=all |
| By: | Arntz, Melanie (ZEW Mannheim); Gregory, Terry (IZA); Zierahn, Ulrich (ZEW Mannheim) |
| Abstract: | Computing power continues to grow at an enormous rate. Simultaneously, more and better data is increasingly available and Machine Learning methods have seen significant breakthroughs in the recent past. All this pushes further the boundary of what machines can do. Nowadays increasingly complex tasks are automatable at a precision which seemed infeasible only few years ago. The examples range from voice and image recognition, playing Go, to self-driving vehicles. Machines are able to perform more and more manual and also cognitive tasks that previously only humans could do. As a result of these developments, some argue that large shares of jobs are “at risk of automation”, spurring public fears of massive job-losses and technological unemployment. This chapter discusses how new digital technologies might affect the labor market in the near future. First, the chapter discusses estimates of automation potentials, showing that many estimates are severely upward biased because they ignore that workers in seemingly automatable occupations already take over hard-to-automate tasks. Secondly, it highlights that these numbers only refer to what theoretically could be automated and that this must not be equated with job-losses or employment effects – a mistake that is done often in the public debate. Thirdly, the chapter develops scenarios on how digitalization is likely to affect the German labor market in the next five years and derives implications for policy makers on how to shape the future of work. Germany is an interesting case to study, as it is a developed country at the technological frontier. In particular, the main challenge will not be the number, but the structure of jobs and the corresponding need for supply side adjustments to meet the shift in demand both within and between occupations and sectors. |
| Keywords: | automation, digitalization, unemployment, inequality |
| JEL: | J23 J31 O33 |
| Date: | 2019–06 |
| URL: | http://d.repec.org/n?u=RePEc:iza:izadps:dp12428&r=all |
| By: | Sumit Agarwal (National University of Singapore); Elvira Sojli (University of New South Wales); Wing Wah Tham (University of New South Wales) |
| Abstract: | Commuting imposes financial, time and emotional cost on the labor force, which increases the cost of supplying labor. Theory suggests a negative or no relation between travel and working time for two reasons: travel time is a cost to supplying labor and commuting frustrates the traveler decreasing productivity. We use a unique dataset that records all commuting trips by public transport (bus and train) over three months in 2013 to study if commuting time affects labor supply decisions in Singapore. We propose a new measure of commuting and working time based on administrative data, which sidesteps issues related to survey data. We document a causal positive relation between commute time and the labor supply decision within individuals. Specifically, we show that a one standard deviation increase in commute time increases working time by 2.6%, controlling for individual, location, and time fixed effects. There are two sources of variation in the elasticity of work time to travel time: across individual and within individual (time variation). While part of the cross-sectional variation may be captured by survey data, the time-variation is completely unexplored. First, we find that the cross-sectional variation depends on whether one engages in a service or manufacturing type of job. This cross-sectional variation might be missed out in survey-based responses due to a different selection process, based say on the proportion of industries in the S&P500. Second, we find that there is very large within individual variation in the elasticity, not based on calendar effects, like day of the week or month. We investigate several potential explanations for this result. We find that in professions where interaction with co-workers and with customers is necessary, i.e. service jobs, disruptions in travelling to work cause a backlog and increase the working hours beyond the original travel delay. These (travel delayed) individuals are not compensated for the time that they put in, in addition to the usual number of working hours. This means that there is a cost shift from employer to employee. Given the recent trend of moving from manufacturing to service-based economies, it is most likely the positive elasticity will increase and become a larger economic burden. |
| Keywords: | Commute time, labor supply, elasticity, task juggling, trains, buses, big data |
| JEL: | D1 J22 J24 M54 |
| URL: | http://d.repec.org/n?u=RePEc:cth:wpaper:gru_2018_015&r=all |
| By: | Arina Wischnewsky; David-Jan Jansen; Matthias Neuenkirch |
| Abstract: | This paper retraces how financial stability considerations interacted with U.S. monetary policy before and during the Great Recession. Using text-mining techniques, we construct indicators for financial stability sentiment expressed during testimonies of four Federal Reserve Chairs at Congressional hearings. Including these text-based measures adds explanatory power to Taylor-rule models. In particular, negative financial stability sentiment coincided with a more accommodative monetary policy stance than implied by standard Taylor-rule factors, even in the decades before the Great Recession. These findings are consistent with a preference for monetary policy reacting to financial instability rather than acting pre-emptively to a perceived build-up of risks. |
| Keywords: | monetary policy, financial stability, Taylor rule, text mining |
| Date: | 2019 |
| URL: | http://d.repec.org/n?u=RePEc:ces:ceswps:_7657&r=all |
| By: | Ilan Noy; Cuong Nguyen; Pooja Patel |
| Abstract: | In 2011, Thailand experienced its worst flood ever. Using repeated waves of the Thai Household Survey, we analyse the flood’s economic impacts. In 2012, households answered a set of questions on the extent of flooding they experienced. We use this self-identified flood exposure, and external exposure indicators from satellite images, to identify both directly affected households and those that were not directly flooded but their communities were (the spillovers). We measure the impact of the disaster on income, expenditure, assets, debt and savings levels, directly, and indirectly on spillover households. We also analyse the flood’s impacts across different socio-economic groups. |
| Keywords: | disaster, flood, Thailand, economic impact |
| JEL: | O12 Q54 |
| Date: | 2019 |
| URL: | http://d.repec.org/n?u=RePEc:ces:ceswps:_7644&r=all |
| By: | Alejandro Micco |
| Abstract: | Digital technologies, robotics, and artificial intelligence substitute tasks performed by labor are bringing back old fears about the impact of technology on labor markets and international trade. The aim of this paper is to provide evidence about the causal effect of automation on the labor market and sectoral US imports. We use robots per workers, instrumented by robot penetration in Europe, to study employment in almost 800 occupations in 285 industries in the US during 2002-2016. We use Autor et al (2003) and Frey and Osborne (2017) methodologies to define occupations at risk of automation and to study their behavior after robots´ penetration. We find that employment in occupations at risk has been declining at an annual rate of 2.0-2.5%, relative to other occupations. This result is mainly driven by a substitution effect within industries defined at the 4-digit NAICS level. One standard deviation increase in robots per worker reduces employment growth by 1.25-1.45% in occupations at risk compared to the other professions in the same sector. Industries with a higher share of occupation at risk have a lower rate of employment growth during the period 2002-2016. Also, imports of commodities produced by these sectors have been falling, in particular from countries with lower penetration of automation technologies. This result suggests that automation is changing countries´ comparative advantage. |
| Date: | 2019–07 |
| URL: | http://d.repec.org/n?u=RePEc:udc:wpaper:wp486&r=all |
| By: | Ho Fai Chan; Bruno S. Frey; Ahmed Skali; Benno Torgler |
| Abstract: | By examining discrepancies between officially reported GDP growth figures and the actual economic growth implied by satellite-based night time light (NTL) density, we investigate whether democracies manipulate officially reported GDP figures, and if so, whether such manipulation pays political dividends. We find that the over-reporting of growth rates does indeed precede increases in popular support, with around a 1% over-statement associated with a 0.5% increase in voter intentions for the incumbent. These results are robust to allowing the elasticity of official GDP statistics to NTL to be country specific, as well as accounting for the quality of governance, and checks and balances on executive power. |
| Keywords: | manipulation, political entrenchment, electoral cycles, trust, popular support, GDP, night lights |
| JEL: | D72 D73 O43 |
| Date: | 2019 |
| URL: | http://d.repec.org/n?u=RePEc:ces:ceswps:_7653&r=all |
| By: | Aufegger, Lisa; Bicknell, Colin; Soane, Emma; Ashrafian, Hutan; Darzi, Ara |
| Abstract: | Background: Small group research in healthcare is important because it deals with interaction and decision-making processes that can help to identify and improve safer patient treatment and care. However, the number of studies is limited due to time- and resource-intensive data processing. The aim of this study was to examine the feasibility of using signal processing and machine learning techniques to understand teamwork and behaviour related to healthcare management and patient safety, and to contribute to literature and research of teamwork in healthcare. Methods: Clinical and non-clinical healthcare professionals organised into 28 teams took part in a video- and audio-recorded role-play exercise that represented a fictional healthcare system, and included the opportunity to discuss and improve healthcare management and patient safety. Group interactions were analysed using the recurrence quantification analysis (RQA; Knight et al., 2016), a signal processing method that examines stability, determinism, and complexity of group interactions. Data were benchmarked against self-reported quality of team participation and social support. Transcripts of group conversations were explored using the topic modelling approach (Blei et al., 2003), a machine learning method that helps to identify emerging themes within large corpora of qualitative data. Results: Groups exhibited stable group interactions that were positively correlated with perceived social support, and negatively correlated with predictive behaviour. Data processing of the qualitative data revealed conversations focused on: (1) the management of patient incidents; (2) the responsibilities among team members; (3) the importance of a good internal team environment; and (4) the hospital culture. Conclusions: This study has shed new light on small group research using signal processing and machine learning methods. Future studies are encouraged to use these methods in the healthcare context, and to conduct further research on how the nature of group interaction and communication processes contribute to the quality of team and task decision-making. |
| JEL: | J50 |
| Date: | 2019–06–13 |
| URL: | http://d.repec.org/n?u=RePEc:ehl:lserod:101073&r=all |
| By: | Arntz, Melanie; Gregory, Terry; Zierahn, Ulrich |
| Abstract: | Computing power continues to grow at an enormous rate. Simultaneously, more and better data is increasingly available and Machine Learning methods have seen significant breakthroughs in the recent past. All this pushes further the boundary of what machines can do. Nowadays increasingly complex tasks are automatable at a precision which seemed infeasible only few years ago. The examples range from voice and image recognition, playing Go, to self-driving vehicles. Machines are able to perform more and more manual and also cognitive tasks that previously only humans could do. As a result of these developments, some argue that large shares of jobs are "at risk of automation", spurring public fears of massive job-losses and technological unemployment. This chapter discusses how new digital technologies might affect the labor market in the near future. First, the chapter discusses estimates of automation potentials, showing that many estimates are severely upward biased because they ignore that workers in seemingly automatable occupations already take over hard-to-automate tasks. Secondly, it highlights that these numbers only refer to what theoretically could be automated and that this must not be equated with job-losses or employment effects - a mistake that is done often in the public debate. Thirdly, the chapter develops scenarios on how digitalization is likely to affect the German labor market in the next five years and derives implications for policy makers on how to shape the future of work. Germany is an interesting case to study, as it is a developed country at the technological frontier. In particular, the main challenge will not be the number, but the structure of jobs and the corresponding need for supply side adjustments to meet the shift in demand both within and between occupations and sectors. |
| Keywords: | Automation,Digitalization,Unemployment,Inequality |
| JEL: | J23 J31 O33 |
| Date: | 2019 |
| URL: | http://d.repec.org/n?u=RePEc:zbw:zewdip:19024&r=all |