IV.4 Discussion in:

Oliver Holtemöller (Ed.)

How Can We Boost Competition in the Services Sector?, page 170 - 176

1. Edition 2017, ISBN print: 978-3-8487-4676-7, ISBN online: 978-3-8452-8902-1,

Bibliographic information
Discussion Dr Stefan Profit: Thank you for both excellent presentations. I would like to make some very short remarks on both even though they would deserve a more pro‐ found discussion, and I will make some general remarks to the topic of this event. Let me start with the paper on the crafts reform: It focuses on the issue of entrepreneurship which has been only one goal of the reform, though an important one. Obviously, the reform fostered and supported self-employment, and consequently we observe a substantial increase in entry probabilities in the aftermath of the reform. On the other hand, we see exit rates being more or less unchanged. Actually, the last finding is a bit controversial and is contradicted by other studies, for instance by Bizer et al. (2014). Hence, there is a discussion whether we see an increase in self-employment or whether the newly created firms were only short lived. The evidence is not very clear. However, the authors find a clear increase in entry probabilities and also in self-employment. If we take a broader picture and not only look at entrepreneurship, there is another recent paper by Koch and Nielsen (2016) which partly summarises the literature but also adds some new evidence using IAB data. On the one hand, they find similar results: they also find higher start-up dynamics. On the other hand, they do not find any effect on aggregate employment, innovation, wages or turnover. Therefore the effects of the crafts reform in Germany seem to be fairly mixed. Firm entries seem to have increased, but we do not ob‐ serve much of an effect on a macroeconomic level. Even on training, where there is a discussion of whether the training effort of companies has decreased throughout the reform, the evidence is not clear. What we see is that training has decreased in restricted occupations, but evidence shows that this only happened after 2009, i.e. after a certain time. So it is quite unclear whether this effect was due to the reform. The result is also sensi‐ tive to the data source. Different studies find different results on the train‐ ing evidence. Mr Palige mentioned a study which looked at prices in crafts and different occupations. I found it hard to find evidence on price devel‐ opments or consumer benefits and I think there is some scope for further IV.4 IV Past Reforms in the Services Sector and their Effects 170 research there because enhancing consumer benefit and demand actually was one of the main goals of the reform. Some technical issues concern the selection of control groups. There has been a discussion on whether you can differentiate the data properly. It is a question of whether you grasp the full effect of the reform within the data that you use. Also, there may be an issue of dealing with unobserved heterogeneity and eliminating it completely by applying the difference in difference approach. There are other econometric approaches like match‐ ing which might be worth exploring. A more general issue which has been mentioned by Henrik Enderlein before and others as well is the link with other reforms. I am not only re‐ ferring to the possibility that empirical evidence on the effects of the crafts reform might partly be blurred by the introduction of labour market pro‐ grammes which foster certain forms of self-employment. The crafts re‐ form was part of a larger reform, the Agenda 2010, which implies syner‐ gies in structural reforms that sometimes cannot be grasped by microeco‐ nomic analyses. If you go to judge the crafts reform, you have to look at the broader approach of reforms on product and labour market in general at that time. On the second paper, I would like to mention a few things as well. I think the design is novel and very innovative. It combines new data, the CompNet data, which is very useful because it draws on micro data from different countries and makes it comparable across countries, with the OECD product market regulation indicators. This is a new approach which is potentially very useful. What you find is the Blanchard and Giavazzi (2003) story ‘Short-term Pain, Long-term Gain’ which is well-known and intuitive, and Henrik En‐ derlein already elaborated on it. One useful aspect of the paper is to look at spillovers across the value chain. Very often when we discuss about reg‐ ulated professions we hear the argument that this sector is not so important from a macro view because it is relatively small. It is something that is of‐ ten underestimated. These sectors are very important e.g. for the manufac‐ turing sector because they deliver inputs for production and through this channel impact the competitiveness of our economy, even though they very often are domestic sectors. One or two technical remarks: There is an issue about endogeneity in structural VARs. You do a Cholesky decomposition and impose a causality chain. This could be discussed. Moreover, one issue, I was not sure about is, whether the fact that you model regulation as an autocorrelated process IV.4 Discussion 171 in part drives your results of ‘Short-term Pain, Long-term Gain’. If you look at your graph with the inverse U-shape path of deregulation: You have minus 20 initially, that is the deregulation reform. But then it raises again and becomes positive. This means you implicitly assume a re-regu‐ lation if you like. Also, even if the CompNet data is very useful it would be desirable to have more disaggregated evidence on firm-level data, which is however difficult to compare across countries. Let me mention one or two more general remarks on what we can learn from past reforms. I think we have been presented two approaches here. One is like a birds’ eye view looking at reforms in a whole sector of trans‐ port, not going into detail of which reforms we actually speak about. The second is a case study that uses a real reform, the German crafts reform from the 2000’s and uses it as a natural experiment to infer evidence on the effects of this reform. Beyond these very useful approaches, we see, in my view, very little research regarding product market reforms that focus‐ es on real welfare effects. Most studies focus on certain aspects like job creation, output, training, wages, firm exit and entry but there is very little evidence on consumer benefits. But, what is the advantage for the con‐ sumer from the crafts reform or from transport deregulation? Is there one at all? Actually, if you look at the recital of the German crafts reform act at the time you find formulations like ‘provide customers a broader scope of services’. So that’s actually the goal, that was written down by the Members of Parliament who decided on that the reform. However, I see very little evidence on that; I am not claiming that the consumer benefitted from it, we just don’t know. There are two issues which were already raised by State Secretary Zy‐ pries. First, I think there is the important issue of quality of services. We did a big share of work, particularly the OECD did, on creating product market indicators. However there is little to no evidence on product quali‐ ty. In my view, we should also have quality indicators across countries to compare them with regulation indicators, to be fair. Second, there is an is‐ sue of resilience of institutions, in particular with respect to global chal‐ lenges like digitization, globalisation, demography and other. For the Ger‐ man case, it is very common to delegate certain regulatory issues to the social partners or professional bodies. You can see this mostly in the labour market. As State Secretary Zypries mentioned, in the 90’s the Ger‐ man model of social partners and industrial relations was considered as one of the most important obstacles to competitiveness. Nowadays it is considered one of the greatest advantages. Therefore, I think you have to IV Past Reforms in the Services Sector and their Effects 172 take a broader perspective when looking at regulation, not only focussing a specific regulation but taking account of the regulatory framework of the economy and the specifically the industrial system as a whole. Thank you very much. Dr Davud Rostam-Afschar: The first point is that the finding of non-increasing or only modestly in‐ creasing exit rates is controversial. My research and other research has found that both entries increased and the self-employment rate increased. Now consider a strong increase in exit rates: the finding that the self-em‐ ployment rate increased cannot be reconciled with such a change. How‐ ever, these higher self-employment rates are what the literature consistent‐ ly finds. This could be a short term effect, therefore it is interesting to look at survival rates, which was beyond the scope of my study. However, I provided evidence on the treatment effect in each of the five years after the reform, which were the latest available data of the German micro cen‐ sus at that time - and still there was a significant treatment effect on en‐ tries and the self-employment rate. I do not think, this is a controversial finding. A second point referred to technical issues. Matching techniques for non-experimental settings have recently developed in a fascinating litera‐ ture. However, the setting in my study is quasi-experimental because treat‐ ment and control groups are defined quite randomly by the law. Therefore other concerns seem more important. Finding a good control group is often difficult and you can debate long, whether you can maintain the common trend assumption. Therefore, I did robustness checks and compared alternative definitions of control groups. For instance, I compared the B1 occupations as treated to the control group of all the A occupations and the results turned out to be similar. A different control group candidate would be the B2 occupations but there is too little information on who works in these occupations. Another alterna‐ tive is to use industry occupations that are similar to crafts occupations but again, it is hard to distinguish these occupations. Taming unobserved heterogeneity may be challenging in many settings, however, in a setting where you compare the same person before and after the reform, the person-specific unobserved heterogeneity would drop out if it is unrelated to the reform. Of course, there are many more extensions IV.4 Discussion 173 that future research could investigate. However, I would be surprised, if the results turn out to be different from what I presented. The third point was to consider the effects of the broad package of the agenda 2010. This is accounted for in the difference-in-difference-ap‐ proach, as I check, for instance, whether a startup subsidy for unemployed or the enlargements to the European Union influenced the results. Beyond that, I leave to others to debate the forces that cannot be captured by mi‐ croeconometric methods. Fourth and finally, the reason that Koch and Nielen (2016) do not find employment effects is that in the IAB data, there are no one person com‐ panies. Their investigation of more indicators is a valuable step towards a welfare analysis. However, for such an analysis you need a fully-fledged structural model, and it is hard enough to measure the effect of one indica‐ tor in an unconfounded way given the available amount and quality of da‐ ta. The contribution of my paper and those who use the same methods like Koch and Nielen (2016) is that future research can use the presented re‐ sults to form a structural model from which one can calculate welfare ef‐ fects on consumers. Key variables that are currently not available are the quantity and quality of consumer services. So how can you say more on the effect of the reform on quantity, quality and prices? As has been men‐ tioned, we need more and better data to make an economic policy possible that is founded in facts and evidence instead of ideological quarrels. If you have some money spare for conducting a survey or an experiment I am happy to offer my services to develop and conduct the research. Thank you very much. Paolo Mengano: Thank you for the comments. Regarding the endogeneity issue you raised, concerning also the Cholesky decomposition and the autocorrelation of the regulation, it is something that I have given a lot of attentions. Thus, after obtaining those results, I ran several tests. I have not found anything that will make me doubt their quality. Moreover, as further robustness check, I re-ran the regulation in first differences, exactly to check whether it might change something, and the results still hold. The Cholesky decomposition, it allows me to get exactly what I wanted to look into, i.e. the exogeneity of the regulation and the desired casual chain among the variables of inter‐ est. IV Past Reforms in the Services Sector and their Effects 174 I tried it with other specifications, too. For example, I tried the same ex‐ ercise with sign restrictions. However, lacking a proper theory, I could do it only for my first result as I could use the restrictions from existing mod‐ els. For the other results, I could not do it because I did not know what restrictions to impose. About the welfare effects that you were mentioning, I believe they are extremely important too. I have not done it so far mainly because this would require more degrees of freedom in my model. Adding them now would likely bias all the estimations. The same stands for prices. More‐ over, I haven’t found a good series of data to add at such a granular level. Thank you very much. IV.4 Discussion 175

Chapter Preview



‘How Can We Boost Competition in the Services Sector?’ is a key question in the process of creating a more effi-cient economic environment in Germany. This book contains a collection of conference contributions on service sector reforms from members of academic institutions, ministries, the EU Commission and other organisations. The conference consisted of a keynote on the importance and implementation of structural reforms in Europe and two panels that dealt with the evaluation of past reforms in the services sector and the potential scope and effects of further reforms.

Since the 1990s, productivity growth in Germany and other Member States of the European Union has been significantly lower than in the US. The development of productivity growth in the services sector is estimated to account for two thirds of this widening gap. The European Commission advocated reforms in the services sector in its country-specific recommendations for Germany. At a conference in Berlin in July 2016, experts from various fields presented and discussed studies on service sector reforms.

With contributions by

Oliver Holtemöller, Brigitte Zypries, Joaquim Nunes de Almeida, Dirk Palige, Henrik Enderlein, Stefan Profit, Davud Rostam-Afschar, Paolo Mengano, Oliver Arentz, Erik Canton, Jochen Andritzky


„Wie können wir den Wettbewerb im Dienstleistungssektor stärken?“ Dies ist eine Schlüsselfrage für eine größere Leistungsfähigkeit des ökonomischen Umfelds in Deutschland. Dieses Buch versammelt Konferenzbeiträge von Mitgliedern wissenschaftlicher Einrichtungen, von Ministerien, der EU-Kommission und anderen Organisationen zu Reformen im Dienstleistungssektor. Die Konferenz umfasste einen Eröffnungsvortrag zur Bedeutung und Durchführung von Strukturreformen in Europa und zwei Gesprächsforen zur Bewertung vergangener Reformen im Dienstleistungssektor und zur möglichen Reichweite sowie zu den möglichen Auswirkungen weiterer Reformen.

Die Zunahme der Produktivität ist seit den 1990er Jahren sowohl in Deutschland als auch in anderen Ländern der Europäischen Union deutlich geringer als in den USA. Es wird geschätzt, dass die Entwicklung des Produktivitätszuwachses im Dienstleistungssektor für zwei Drittel dieses zunehmenden Abstandes verantwortlich ist. Die Europäische Kommission spricht sich in ihren länderspezifischen Empfehlungen zu Deutschland für Reformen in diesem Sektor aus. Auf einer Konferenz im Juli 2016 in Berlin stellten Experten aus unterschiedlichen Bereichen Studien zu solchen Reformen vor und diskutierten deren Ergebnisse.

Mit Beiträgen von

Oliver Holtemöller, Brigitte Zypries, Joaquim Nunes de Almeida, Dirk Palige, Henrik Enderlein, Stefan Profit, Davud Rostam-Afschar, Paolo Mengano, Oliver Arentz, Erik Canton, Jochen Andritzky