What the Mafia teaches about fiscal policy

What’s the impact of fiscal policy on GDP growth? That’s one of the most contentious questions in economics. A lot of economists regard economic stimulus packages as completely useless. New empirical evidence shows that they are wrong. Amazingly, we know this partly thanks to the Italian mafia.

Salvatore Riina

Salvatore Riina: The most powerful mafia boss of the 1980s, after his arrest in 1993 (image via Wikipedia)

It was one of the most expensive signatures in American history. Barack Obama had flown to Denver, Colorado, to give it. On 17 February 2009, the American president signed the “American Recovery and Reinvestment Act“, an economic stimulus package worth $787 bn. It was the biggest single piece of fiscal policy in the history of mankind.

Governments all across the globe acted accordingly. In their fight against the most severe economic crisis for decades they spent billions on fiscal policy.

Did this debt-financed policy avert a second Great Depression? Or was it just massively expensive and futile?

The answers economists are giving to these questions vary a lot. For several decades the mainstream consensus has been that fiscal policy just does not work. Recently, however, the views in the profession have begun to change. There is growing empirical evidence that fiscal policy is better than its reputation.

“The last decade has produced a substantial improvement in economists’ understanding of the empirical orders of magnitude of a number of fiscal impacts,” wrote Norman Gemmell recently in the “Economic Journal” (“Can a Fiscal Stimulus Boost Economic Growth Rates?“).

Gemmell is the  Chief Economist of the New Zealand Treasury and a former Professor of Economics of the University of Nottingham. “Consensus”, Gemmell stressed, “could not yet be said to have been reached.”

What seems to be clear, however, is that the complete dismissal of fiscal policy is based on few empirical facts.

Theoretically, it’s quite clear cut…

Theoretical considerations suggest that government purchases crowd out private activities. If humans were fully rational and if all markets worked perfectly, we would smooth our lifetime consumption. Temporary income gains would not influence our current spending. We would anticipate that debt-financed tax cuts imply higher taxes in the future.

Hence, we would not spend the extra money but save it. Therefore, apart from the higher public deficit, such tax cuts would not have any impact on the economy.

Economists call this phenomenon “ricardian equivalence”. It is a unifying feature of the most modern macro models. It is, therefore, small wonder that theoretical simulations based on these models come to the conclusion that the stimulus packages won’t do much good. (see: John Cogan, Tobias Cwik, John Taylor and Volker Wieland: “New Keynesian versus Old Keynesian Government Spending Multipliers”)

How the Mafia helps economists

However, real life seems to be different. According to a bunch of empirical papers, “fiscal injections can boost output and consumption levels in the short-run”, writes Norman Gemmell.

United States President Barack Obama signs int...

Expensive moment: Barack Obama signs the “American Recovery and Reinvestment Act" (image via Wikipedia)

The scientific evaluation of fiscal policy is quite a task. The mere observation that the world economy has recovered quickly from the recession of 2009 does not say anything. Nobody knows how things would have panned out without the economic stimulus packages. Economists also often struggle to disentangle cause and effect: Does growth accelerate because government spending increases, or is it the other way round?

Thanks to the Italian mafia, three economists have recently been able to solve this problem. In a fascinating paper entitled “Mafia and Public Spending: Evidence on the Fiscal Multiplier from a Quasi-experiment”, Antonio Acconcia, Giancarlo Corsetti and Saverio Simonelli analysed the effects of fiscal policy inItaly. They did not look at the whole country but focused on small regions.

Organised crime is a nightmare inItalybut it is a blessing for economists because it causes significant changes in government expenditures that are completely unrelated to the current economic environment. If a politician is arrested because of his connections to the mafia, all public investments in infrastructure and construction are put on hold immediately until the investigators have checked if the mafia had been able to channel money into their pockets.

“Public work and projects are started again only after investigation and scrutiny of previous tender procedures and decisions to establish that the contractors are not effectively mafia Associations,” the economists explain. “The dismissal of elected administrators is thus associated with sizeable fiscal retrenchment.”

Acconcia, Corsetti and Simonelli looked at more than 100 cities where local politicians were forced to resign because they were Mafia figureheads. Afterwards, local government expenditures declined on average by 20 percent.

It’s all about the multiplier

According to the paper, this was highly detrimental for the local economy. When the government expenditures were slashed by one Euro, the economic activity immediately shrank 1.40 Euro. In the medium term, the output losses accumulated to 2 Euro. Hence, the so called government multiplier was between 1.4 and 2.

These estimates are in line with a number of empirical studies on the macro level. In a very influential paper, Olivier Blanchard and Roberto Perotti in 2002 analysed American fiscal policy after 1947. Their paper had a rather clumsy title (“An Empirical Characterization Of The Dynamic Effects Of Changes In Government Spending And Taxes On Output”) and was published in the “Quarterly Journal of Economics”.

Their conclusion was as follows:

“We consistently find a positive effect of government spending on private consumption, a straightforward implication of virtually all Keynesian models but a result that is difficult to reconcile with the neoclassical approach.”

According to Blanchard and Perotti, the fiscal multipliers were rather small, often close to one.

Five years later, other researchers published estimates that were even higher. In their paper “Understanding the Effects of Government Spending on Consumption,” Jordi Gali, Javier Vallés and J. David López-Salido came to the conclusion that one Dollar of higher government spending in the U.S. increases GPD by 1.74 Dollars after two years. Private consumption had risen by 0.95 Dollar. In a different paper, Jordi Gali concluded:

“Keynes and his followers got it right.”

John Maynard Keynes: Inventor of fiscal policy (Image via Wikipedia)

Roel Beetsma and Massimo Giuliodori, two economists with the Universityof Amsterdam,  have recently published similar estimates of the impact of fiscal policy for Europe. With regard to EU member states , they came to the conclusion that a 1% increase in government expenditure leads to a GDP increase of around 1.5% after one year.

“The multiplier is quite large, but not out of line with findings by others,” wrote Beetsma and Giuliodori in their paper entitled “The Effects of Government Purchases Shocks: Review and Estimates for the EU.”

The devil is in the details

However, the effects of fiscal policy vary quite significantly from country to country. The more internationally interconnected a country is, the smaller the effects of fiscal policy, the economists observe. In open economies, the fiscal stimulus partly leaks overseas.

Similar results are obtained by Ethan Ilzetzki (London School of Economics), Enrique Mendoza and Carlos Végh (both: Universityof Maryland). In their paper entitled “How Big (Small?) are Fiscal Multipliers?” they conclude:

“The medium to long-run effects of increases in government consumption vary considerably. In particular, in economies closed to trade or operating under fixed exchange rates we find a substantial long-run effect of government consumption on economic activity. In contrast, in economies open to trade or operating under flexible exchange rates, a fiscal expansion leads to no significant output gains.”

A lot hinges on the nitty gritty details of the stimulus packages. For example, “cash for clunkers programmes” seem to be egregiously inefficient, Atif Mian and Amir Sufi show in a paper entitled “The Effects of Fiscal Stimulus: Evidence from the 2009 ‘Cash for Clunkers’ Program.”  They conclude:

“Our results reveal a swift reversal in auto purchases at the expiration of [the 2009 Cars Allowance Rebate System] which highlights a strong inter-temporal substitution that quickly “crowds out” the initial effect. Our evidence suggests that the ‘cash for clunkers’ program, a program that cost $2.85 billion, had no long run effect on auto purchases.”

Handle with care

Hence, fiscal policy is no golden bullet and should be used with thorough consideration by governments. Probably deficit spending should only be done in very rare and extraordinary occasions: when severe recessions threaten economic stability. Even though economic stimulus packages are able to thwart a total economic collapse in the short term, from a medium to long term perspective the assessment is different, the economists Norman Gemmell, Richard Kneller and Ismael Sanz conclude in a paper entitled “The Timing and Persistence of Fiscal Policy Impacts on Growth: Evidence from OECD Countries”:

“We find that positive growth effects associated with ‘productive’ public spending changes have often been approximately counteracted by tax changes with negative effects. Furthermore, growth-affecting fiscal policy changes are often reversed, so growth effects in practice are expected to be short-lived.”


If you happen to like this blog, why don’t you like me on Facebook as well?


Filed under Economic Policy, Financial Crisis, General Economics

25 Responses to What the Mafia teaches about fiscal policy

  1. Pingback: » What the Mafia can teach us about the fiscal spending multiplier Economics and Mechanisms

  2. Pingback: Insights on Stimulus, Thanks to the Mafia - NYTimes.com

  3. Pingback: Economix: Insights on Stimulus, Thanks to the Mafia – New York Times (blog) « Uk Corruption « CORRUPTION!

  4. Pingback: Bin Hammam says reputation under Blatter is ‘sullied’ – TSN « Uk Corruption « CORRUPTION!

  5. Pingback: Economix: Insights on Stimulus, Thanks to the Mafia | G7Finance.com - Finance News & Personal Finance Resources

  6. Pingback: Secondary Sources: Mafia and Stimulus, Housing and Recovery, Trade « Finance Blog

  7. Pingback: Secondary Sources: Mafia and Stimulus, Housing and Recovery, Trade - Real Time Economics - WSJ

  8. Pingback: Economix: Insights on Stimulus From the Mafia | CERES ATACADO

  9. Pingback: Economix: Insights on Stimulus From the Mafia

  10. Pingback: Italian economists test fiscal stimulus with the help of mafia

  11. Pingback: What the Mafia teaches about fiscal policy | Economics Intelligence | Articles about the world

  12. Pingback: Financial news » Blog Archive » Economix: Insights on Stimulus From the Mafia

  13. Pingback: The Marginal Prophet | Lee Geiger » Blog Archive » Nelson Goes to Ecuador

  14. Could it be, that the mafia in some regions ‘makes the world go round’ and that there occurs an economic standstill after their arrest for this reason, rather than the direct effect of the decrease in public spending?

  15. Very interesting piece, but something is unclear to me:

    “Acconcia, Corsetti and Simonelli looked at more than 100 cities where local politicians were forced to resign because they were Mafia figureheads. Afterwards, local government expenditures declined on average by 20 percent.”

    By ‘afterwards’ do you mean after the mafia influence has been removed – or only while spending is put on hold during investigations?

    Regardless of the answer, the Italy’s mafia has been receiving handouts from cronies in local government.

    As an example of just how extensive the mafia/local government relationship in Italy is, two stories: one of a councillor in Liguria being arrested for mafia collusion: http://www.ilfattoquotidiano.it/2011/05/12/corruzione-e-lombra-della-ndrangheta-in-carcere-il-candidato-del-pd-a-savona/110674/

    And another example is that of one of the candidates standing for election in Milan’s administrative elections – who was caught chatting about killing someone with a mafia boss. The candidate has not withdrawn from the elections and is still listed as a candidate here: http://www.comunalimilano2011.it/users/viewprofile/501/elections

    I’ve written a little about this: http://italychronicles.com/milan-test-for-berlusconi/

    Mafia infiltration into politics in Italy at all levels is widespread, so widespread that not enough seems to be done about it.

    Odd place, is Italy.

    All the best from Milan,


  16. Pingback: The Marginal Prophet- Nelson Goes to Ecuador | The Marginal Prophet

  17. Pingback: Friday Reads | The Big Picture

  18. At first glance, this pseudo-experiment seems to overstate the multiplier by approximately 1, because the decreased spending in the affected localities was not accompanied by decreased taxation or borrowing. (A naive Keynesian model might assume that borrowing has no effect on the multiplier, while a rational-expectations model would lead to the conclusion that taxation and borrowing are nearly equivalent.)

    This omission would seem to reduce the expected multiplier for national stimulus spending — which must be paid for by the nation receiving the stimulus — to between 0.4 and 1.0.

  19. Pingback: Friday Reads : Invest My Money

  20. Jacob, Alex and Sammler, thank you all for your comments. I really appreciate your feedback.

    @ Jacob: This is an important issue. I had another look at the paper. On page 11, the authors are at least partly addressing this point. They write:

    “Arguably, a reason why this condition [that the dismissal of criminal councils affects GDP only through government spending] might fail is that, by calling attention to mafia activities in the local district, the dismissal of the city council may have economic consequences independently of the cut in public spending. For instance, the mafia may relocate some of its business for fear that the whole area will be subject to intense police investigation. If this is the case, however, it is plausible to expect that areas with a relatively heavy presence of the mafia would also be characterized by a relatively high number of mobster arrests by the police. Under this maintained hypothesis, we control for the direct channel of city council dismissals on the economy (independent of the multiplier effects) by including the number of people reported to the judicial authority for organized crime, extortion, mafia-related murders and corruption, and also the number of corruption crimes, in our set of regressors.”

    This is probably not exactly what you have in mind, Jacob. I’ll get in touch with the authors and ask them what their answer to your question is. I would suspect that you’re overestimating the effect of the arrests. We are “only” talking about arrests of politicians with connections to the mafia – not about major crackdowns on the mafia itself (for example: arrests of godfathers and dozens of mobsters, seizure of wealth). Hence I don’t think that the issue you’re raising drives the results.

    @ Alex: Those stories about Berlusconi and his subordinates are really hair-rising…

    Anyway, “afterwards” relates to the investigation period. On page 4 of their paper the economists write “the average growth rate of spending at provincial level turns negative conditional on a municipality being placed under compulsory administration, with an average contraction of 20 percentage points.”

    @ Sammler: Intersting point. However, I’m not sure if I go along. That’s for several reasons.

    1) The fist one is a mere technicality. When economists talk about the spending multiplier, they always only look at the spending, not at the financing. Wikipedia puts it like this: “The fiscal multiplier is the ratio of a change in national income to the change in government spending that causes it.” The question always is: “What happens to the GDP this year / next year / the following year when government expenditures change by x Dollars?”

    Hence, future tax increases which might be caused by deficit spending are only taken into account if they affect GDP growth in that period (for example, if taxpayers are forward looking and increase their savings because they want to be prepared for higher taxes in the future, an issue that is discussed in the blog).

    As far as I understand the literature, the possible distortions on the allocation of labour or capital that might come with higher taxes in the future – the deadweight loss of taxation – are not taken into account when the simple spending multiplier is discussed.

    2) One might argue that only a gross multiplier that includes all dynamic effects including the deadweight loss of future taxes makes sense. Actually, there are some papers who try to do this. One is by Gemmell, Kneller and Sanz and it is briefly discussed at the end of my post. They come to the conclusion that “positive growth effects associated with ‘productive’ public spending changes have often been approximately counteracted by tax changes with negative effect”.)

    However, your calculations rests on the implicit assumption that lower taxes or less public borrowing automatically transforms into GDP generated by the private sector.

    Well, if this was the case, the whole discussion about fiscal policy and the multiplier could stop.

    This might be true according to theoretical models based on rational expectations. Cogan/Cwik/Taylor/Wieland make exactly this point in their 2009 paper I refer to in the blog post. However, the whole blog post is about the fact that the empirical observations are apparently not in line with those theoretical models.

  21. Pingback: What The Mafia Teaches About Fiscal Policy — Catchy title, serious issue: does economic stimulus work? It can be difficult to tell because when policymakers do it they don’t know what would have happened if they hadn’t. Happily, Italy&am

  22. Pingback: Insights on Stimulus, Thanks to the Mafia | WebAdvicer | Economy

  23. Pingback: One year on – Top 10 posts on “Economics Intelligence” | Economics Intelligence

  24. Pingback: Sinnlose Schätzungen des Multiplikators? | Kantoos Economics

  25. Pingback: A lot of economists regard economic stimulus packages as completely useless. New empirical evidence shows that they are wrong. Amazingly, we know this partly thanks to the Italian Mafia. « Economics Info