Discussion Paper

No. 2016-31 | June 30, 2016
Is Financial VAT Neutral to Financial Sector Size?
(Published in Special Issue Recent Developments in Applied Economics)


The influence of the taxation of financial services in VAT on financial sector size is analyzed empirically. The authors use data from 36 countries of the European Union and the OECD for the period between 1961 and 2012. Dynamic panel data techniques are used, concretely the GMM System. An unbalanced panel is handled. The results allow them to support the theoretical analysis that states financial VAT has no significant effect on financial sector development, being neutral in relation to this variable. Results are robust to the specifications of the dependent variable.

JEL Classification:

H25, H21, E62, E44, G21


  • Downloads: 1312


Cite As

Julio López-Laborda and Guillermo Peña (2016). Is Financial VAT Neutral to Financial Sector Size? Economics Discussion Papers, No 2016-31, Kiel Institute for the World Economy. http://www.economics-ejournal.org/economics/discussionpapers/2016-31

Comments and Questions

José M. Domínguez - Invited Reader Comment
August 01, 2016 - 08:56

see attached file

Julio López-Laborda and Guillermo Peña - Reply to Comment
September 06, 2016 - 11:52

see attached file

Anonymous - Referee Report 1
August 05, 2016 - 09:36

This is a very good paper, excellent indeed. The motivation, literature review and econometric analysis are very thorough. The paper has important policy implications, which make this piece of research a good contribution to the literature.

However, the paper is an empirical exercise (whether financial VAT is neutral to ...[more]

... firm size). Robustness is essential for empirical papers. So, authors should made an attempt to look for robustness by applying alternative methods and looking for the potential existence of asymmetries in order to avoid bias. Below some concerns are expressed in terms of questions regarding methods and data used in the research. These questions should be dealt with implicitly or explicitly in the text of the paper. I advise a major revision addressing these issues.
A table summarizing previous literature in terms of econometric methods used and results obtained could be a good thing.

Julio López-Laborda and Guillermo Peña - Reply to Referee Report 1
September 06, 2016 - 11:54

see attached file

Anonymous - Invited Reader Comment
August 10, 2016 - 08:17

The paper "Is Financial VAT Neutral to Financial Sector Size?" by López-Laborda and Peña analyses a panel data set of 36 developed countries to explain a country's financial sector size. A primary concern is the influence of valued added taxation of banking services. The authors conclude that neither a financial ...[more]

... VAT nor other taxes imposed on financial services affect the size of the financial sector in an statistically significant way.

The paper contributes to both the political and the academic debate. The financial crisis has reinforced the notion of an excessively large banking sector and the paper analyses empirically whether or not a financial VAT could help to reduce financial sector size.

Overall, the paper poses a relevant question and its empirical approach nicely complements the theoretical research on the topic. Still, there is room for improving the paper. I have listed some comments and suggestions below.

1. The theoretical results from Aigner and Bierbrauer (2015) are ambiguous: if a financial VAT is deductible, it has no impact on the equilibrium allocation. If a financial VAT is not deductible, its level does impact the equilibrium allocation. López-Laborda and Peña discuss this issue in Section 3 of their paper. But their discussion (for me) was hard to follow. Later in their paper, the authors state that, guided by theory, one should expect that the predictor "financial VAT (fVAT)" has no impact on financial sector size (Table 2). The predictor "fVAT" is 1 if a country applies any sort of taxing financial services and is 0 otherwise (Table 1). In my view, the paper lacks a discussion of the various methods of financial VAT listed in Table 1, though. It is not obvious that we should expect all of these methods to have the same effect on financial sector size, but the paper implicitly presumes this. I am not saying that this presumption is wrong, but the paper in its current form does not convince me of the opposite and the theory, as said before, is ambiguous.

2. The predictor "fVAT" is modeled as a dummy variable. This seems to be a valid first pass, but it would be nice to account for the various tax levels as well. A low financial VAT might be insignificant while a high one could have an effect. The current specification could be misleading in such a cases.
The same comment holds for the predictor "separate".

3. The Introduction does not tell the reader why the topic of the paper is interesting. Section 3 provides some motivation for why the size of the financial sector is an interesting and relevant topic to study. My suggestion would be to integrate this discussion into the Introduction.

4. The paper could benefit from some rewriting of the critical passages. It might be my personal impression, but I had trouble following some of the arguments and descriptions in the paper. In particular:

4a. The first paragraph of Section 3 could benefit from some rewriting to be presents its arguments more clearly.

4b. Pages 5 and 6 discuss the theoretical results from Aigner and Bierbrauer (2015) and differentiate between various cases and scenarios. The section would benefit from some rewriting. In particular, it is not always clear (to me) which result follows form which particular specification.

Julio López-Laborda and Guillermo Peña - Reply to Comment
September 06, 2016 - 11:56

see attached file

Anonymous - Referee Report 2
September 08, 2016 - 08:27

In my opinion the article is attractive and represents an interesting approach to a relevant topic. The contribution therefore is valuable, although I think the manuscript requires MAJOR REVISIONS before it can be reconsidered for publication:

Major comments:

1) The introduction is too short, while the sections 2,3 ...[more]

... and 4 are too long. I think it would be necessary to modify the structure of the four sections. The introduction should be much more attractive and informative. The other three sections should reduce their extension since their contents could appear in the Introduction.

2) The residuals are not white noise (p.19). It should explain whether this was taken into account and the effects it has on the tests performed.

3) Equation 3 contains an error or is poorly explained. T (target vector of variables) should appear transposed. Is there a lack of subscripts?

4) The gamma coefficient is very close to one. Is it possible that the dependent variables present a unit root problem? How does it affect the results?

Minor comments:

1) Table 1 is confusing or incomplete.

2) Table 3 does not appear complete. In any case I think that this table is not necessary.

3) Table 6 contains an error. The variable "t-1 pesosb" should be called "bsize t-1".

4) The tables should include a number of decimal places homogeneous at least in each of the columns.

5) On page 18 (line 3) and on page 19 (lines 10,12), replace "residues" by residuals.

Julio López-Laborda and Guillermo Peña - Reply to Referee Report 2
September 20, 2016 - 08:28

see attached file