Monday 15 August 2011

The Ministry of Finance report on fiscal devaluation

The Portuguese Ministry of Finance has posted a report on the implementation of a fiscal devaluation in Portugal.

I wish to make a few comments:

1. The authors should state that the report is preliminary. The report does not contain any conclusion, therefore, in its current form, it is still incomplete and preliminary.

2. The report is acknowledged to be a patchwork of views of four different institutions (Bank of Portugal, Ministry of Finance, Ministry of the Economy and Employment, Ministry of Solidarity and Social Security). While a list of all four different stances is a useful starting point, the final document must find a coherent and unique position on the matter. After all there is only one entity that must fiscally devaluate, Portugal.

3. The structure of the document appears to be biased against the fiscal devaluation.

3.1. The first quarter of the study consists in a detailed description of the social security's financing sources. It is useful to know these details, and there is a lot of interesting information (see below Remark in 3.3). It is important to know which laws should be abrogated and/or changed if the government implements the reduction of the social security contributions paid by employers. The chapter ends with a comparison between Portugal and EU's fiscal revenues structures and states that Portugal relies more heavily than the rest of EU on indirect taxation (read VAT) and less on social contributions (paid by employers). These numbers are obviously influenced by the composition of aggregate demand and income shares. For example, in Germany private consumption (basically the tax basis for VAT) is around 56% of GDP and in the Netherlands private consumption is 45% of GDP. In Portugal private consumption is around 64% of GDP. I would imagine this section as an appendix and not as the initial chapter of the report.

3.2 The second section which addresses the macroeconomic effects (the core of the matter) of the fiscal devaluation is short. The evaluation is performed using a dynamic stochastic general equilibrium (DSGE) model developed at the Bank of Portugal. Two similar DSGE models developed by the ECB and the European Commission have also been used to confirm the results. Personally I believe that these models are very useful for theoretical guidance but should be complemented by more empirical approaches for a more robust quantitative evaluation of the suggested policy. [For the more analytical inclined readers : in the technical description of the macroeconomic effects, a neat identification of short term effects (fiscal devaluation), medium term effects (firm entry in the tradable sector) and long term effects (new steady state due to possible non neutral shift in aggregate labor demand and labor supply) is absent, which makes the quantitative results difficult to interpret].

3.3 The third section is the central part of the report and describes a menu of options to implement the measure. The benchmark measure, the only one that can be labeled as a fiscal devaluation, is the reduction of the payroll tax across all sectors compensated by an increase in VAT (and/or a decrease in public expenditure). The report correctly identifies the main weakness of the measure in the market power of the non-tradable sector (Remark: I noticed that in Table 3 Section1, a large non-tradable industry, or network industry, such as Telecommunication pays a payroll tax of 7.8%, as opposed to the general rate of 23.7%. Given that such a company-industry will not see its payroll tax being reduced the above weakness does not apply).

The alternatives suggested are not fiscal devaluations but could be labeled as :
a) Incentives to job creation (low payroll tax only for new jobs) ;
b) Export oriented Industrial policy (payroll reduction in export industries);
c) Tradable (?) oriented industrial policy (reduction of payroll tax only for lower wages).

3.4 The last page and a half is concerned with the financing of the payroll tax reduction. The last sentence briefly mentions that if the measure is implemented, the necessary fiscal revenues will have to be found by increasing the lower VAT rates. Obviously the financing of the reduction in the payroll tax is a (the) key aspect for the implementability of the measure. After all, who would disagree to lower labor costs if the sustainability of the social security system was secured? This is the section I would have expected the government to put more (all) efforts in order to reach an educated opinion on how much revenues an increase in the lower VAT rates could generate (0.5% of GDP, 1% of GDP, 2% of GDP…4% of GDP?).

As a final remark: I am confident international observers would welcome a report in English (only the executive summary and a technical appendix by the Bank of Portugal are in English), especially if the government plans to convince the troika not to implement the measure or to transform it in a different policy.


  1. "The structure of the document appears to be biased against the fiscal devaluation".
    Why? Because the findings does not match with your opinions? Or because the figures does not support your thesis?
    I agree this is a key issue for our economy and it's very important to promote a public debate. In a previous post a reader raised very good questions that you have not reply yet. That's not my idea of a public debate.

  2. @ J. Tavares de Moura

    I have explained (maybe not satisfactorily) that the key question left unanswered for the implementation of the fiscal devaluation is "how much revenue can be generated by a restructuring of the intermediate VAT rates". The report spends one sentence on this question. If you let this question unanswered, the fiscal devaluation cannot be implemented. And yes, this is my personal view.

    Regarding the questions I did not answer, would you be so kind to send me a link?


  3. Daniel Ribeiro left some comments on your post of June,28, here (
    One more question: I did not find any assessment of the foreseeable increasing tax evasion with the proposed rise of IVA rates, why?

  4. Before all, I think it is to be congratulated some intelligent remarks made in the post.

    I would like to make a couple of questions.

    1st: Why are there so many different rates. I think I would understand an effort to subsidize/tax some industries in favor of others or even some type of workers. But there are some cases which I can't understand due to their specificity. Looking at that table I see the usual comment that Portuguese tax system is quite confuse and complex.

    2nd: How will the reduction in payroll tax work? Let's say they adopt a 4 percentage points decrease. Will it be applied to all the categories that can be seen in table 3 (even if they left out the unbelievable low rate for PT workers)?

    3rd: Both Prime Minister and the Ministry of Finance stated a reduction of around 9% to 10% in primary expenditures for the year of 2012. A reduction in the payroll tax would have a direct impact on the cost of public sector workers, wouldn't it? If I'm right (which I would like to have a confirmation from a reader) and assuming a 4 pp reduction, a large share of the 9 to 10% reduction would be immediately accomplished. This seems to me to be the only rational argument to apply the tax cut for the entire economy and not only the tradable industries.

    4th: I've been reading that a tax cut for the entire economy will not help improving the foreign account position. Perhaps I'm missing something (and, once again, I would like for someone to tell what if I'm really missing something, which is quite likely as I don't work in these fields) but a reduction in the cost of production of a non-tradable industry will promote a substitution of imports in favor of national production, isn't it? I can imagine that such effect might be negligible but there is some, am I correct?

  5. Fiscal devaluation can eliminate not only payroll tax but also pay-as-you-go personal income tax. Employers can offset both against VAT. See "Fiscal devaluation on steroids".