Attend Spring Meetings on development topics from Apr 18-23. Comment and engage with experts. Calendar of Events

Syndicate content

Multipliers in Europe and Africa

Shanta Devarajan's picture

IMF Chief Economist Olivier Blanchard created quite a stir at the recent American Economics Association Meetings when he presented his joint paper with Daniel Leigh that showed that, for 26 European countries, the fiscal multipliers—the amount by which output expands with an increase in the fiscal deficit—were considerably higher than previously thought.  Whereas these multipliers were previously thought to be around 0.5, they find them to be above 1.0.  Applying these figures to a reduction in the fiscal deficit (sometimes called “fiscal consolidation”), Olivier and Daniel suggest that people may have underestimated the extent to which European economies would contract in the wake of their fiscal consolidation.

Several people have been asking whether the same reasoning applies to African countries, whether expanding the fiscal deficit will lead to a more-than-one-for-one increase in GDP. Or, conversely, whether contracting the deficit by, say, one percent of GDP will lead to a fall in GDP of more than one percent. The short answer is no.  A recent paper by my colleague Aart Kraay estimates multipliers for a sample of 29 low-income, aid-dependent countries (all but four of which are in Africa).  Using a clever econometric technique (that exploits data from World Bank projects), he finds multipliers that are not statistically significantly different from zero.  In other words, increases in the deficit appear not to have an effect on GDP in the short run.

On reflection, this result is not too surprising. An increase in the fiscal deficit (by, for example, an increase in government spending) will stimulate output if that additional spending increases employment of idle resources.  For example, if the government spends on building a road, it will hire workers and purchase construction materials.  If those workers were previously not working, and the construction materials could be produced using existing machinery, then the spending will contribute to an increase in GDP.  However, if those workers were already working elsewhere (perhaps in the informal sector), and the machines are producing at full capacity, then the stimulative effect will be minimal. 

Furthermore, in many countries, a decision by the government to spend on a road does not always lead to a road being built that year—or ever, sometimes.  This is what many observers, including some African policymakers, call “the execution deficit.”  It may also explain why, in the wake of the global economic crisis of 2008-9, many African countries, even though they had fiscal space, ran fairly modest fiscal deficits to cushion their economies from the global recession. 

When I suggested that he might want to increase the fiscal deficit because the global economy was contracting, one finance minister said to me, “Shanta, if I could stimulate the economy by running a higher fiscal deficit, I would have done it last year.”


Submitted by Kako NUBUKPO on
Bonjour Shanta, Je crois que tu as raison, et en plus des raisons que tu évoques, il y a des raisons additionnelles. En fait, l'offre intérieure est faible dans les pays africains, ce qui fait que toute stimulation de la demande se traduit par une hausse des importations. Ceci est valable non seulement pour la demande des biens de consommation, mais également pour les biens d'équipement. La preuve, au Togo où on investit beaucoup actuellement pour construire des routes, on dégrade le solde courant de la balance des paiements car on importe beaucoup de biens d'équipements et de matériaux pour la construction des routes. Le multiplicateur des dépenses publiques est d'autant plus faible que la propension marginale à importer est élevé. Par ailleurs, dans la zone UEMOA, le taux de pression fiscale est de 17% en moyenne, contre 45% en France. Ceci veut dire que toute politique expansionniste des dépenses publiques se traduit par une hausse du déficit budgétaire car les rentrées fiscales sont très faibles, engendrant encore une fois les "déficits jumeaux". Tout ceci est préoccupant car nos pays ont besoin d'investir pour se développer. Alors que faire? Je crois qu'il faut confier aux institutions régionales (Commission de l'UEMOA, de la CEDEAO, de la CEMAC...) le soin d'émettre des titres sur le marché financier régional qui est surliquide, ce qui leur permettrait de mobiliser des ressources pour financer les grands travaux d'infrastructures dont les pays ont tant besoin. Ceci permettrait aux Etats subsahariens de moins s'endetter pour les infrastructures, et surtout obligeraient les institutions régionales à jouer leur rôle de mutualisation des risques et de génératrices de solidarités. Il faudrait que le Banque Mondiale développe plus d'outils d'appui aux institutions régionales, afin qu'elles s'engagent de façon efficace dans les investissements productifs, gage de hausse de la productivité. Keynes n'est pas mort, il change juste d'échelle!

Submitted by Calle on
A small comment on: "However, if those workers were already working elsewhere (perhaps in the informal sector), and the machines are producing at full capacity, then the stimulative effect will be minimal. " But if resources allocated in the informal sector and post stimulation were to be used in the formal sector, then we have an actual (and verifiable) increase in GDP? /Calle

Thanks for your comment, Calle. If these workers get regular, formal-sector jobs (with corresponding wage security), then there will be an increase in GDP. But the fact is that with stimulus-induced programs, the workers who build the roads, etc. are getting paid about the same (or only slightly more) than they earn in the informal sector. Furthermore, these jobs are not permanent. When the road is built, they go back to their informal jobs. Shanta

I have a lay opinion on why the deficits in Africa have near-nil multiplier effects: urban capture. We spend all our money in established, self-sustaining urban economies. If the spending was rural oriented, the impacts would be greater. As we stand, all the money simply follows the same set of factors -urban land, urban capital, urban skilled labour... that yields inflation since more money follows the same acre of land in the CBD year in, year out. I am not surprised, because all that capital is already overinvested. The trick is in getting the money out of old cities, into human services. Save the 25-55 year olds who have the skills and professional capacity to access, obtain and invest credit. As our parents and wealth-creators die off to preventable diseases, we keep raising a broke, broken and orphaned next generation of poorly educated urban immigrants that are essentially unproductive, expensive to sustain and easily prone to chaos and dysfunction. The multiplier effects will never be realized unless we all commit to real transformative investment in the rural populations...

Submitted by Nachiket Mor on
The findings of Kraay (2010) resonate with me. In the Indian context I worry that schemes such as the NREGS have resurrected the old Keynesian multiplier models to build their case for more government spending and higher fiscal deficits. There is no question in India there is a great deal of disguised unemployment and there is urgent need for job-creating growth. However in the presence of a strong income effect I fear that schemes such as the NREGS have merely resulted in a contraction of labour supply and slowed down the process of urbanisation that is so essential to long-term Indian prosperity. This combined with a strong "execution deficit" and very poor project selection mechanisms I fear may be resulting in negative multipliers in India. This is certainly not my area of expertise and I would be eager to learn from you and the comments of your readers if my fears are well founded or not.

Submitted by Nachiket Mor on
I also want to get some guidance on the whole issue of Execution Deficit that is mentioned in your post. In many schemes the government is moving strongly to roll back embedded subsidies and moving them into an Electronic Benefits Transfer mode directly to individuals. There is a lot of debate on it in India but I for one find that I am supportive of this broad approach particularly where the presence of these subsidies has distorted entire markets and led to massive misallocations of resources. In this broad remit I would include schemes such as NREGS as well. However, in some sectors such as elementary education and healthcare it is not at all clear to me if cash or cash-like instruments (school-vouchers, insurance) are really a good substitute for direct provision by the government because of the specific nature of these sectors. And, in countries like India we are battling two Execution Deficits -- one of regulation and the other of provision. The use of cash-like instruments in elementary education and healthcare presumes that we have overcome the Execution Deficit as far as regulation is concerned. I worry that this not the case and is a deficit that is much harder to overcome and that we are better off staying with government provision and working hard internally to overcome the Execution Deficit on the provision end. Once again it would be wonderful to get some guidance from you and your readers on these issues.

Nachiket: Thanks for your thoughtful questions. On cash transfers or vouchers for education (and health), your skepticism is warranted. The fact is that someone still needs to supply education and health services for the recipients of these transfers to benefit. The success of the famous conditional cash transfer program in Mexico (Progresa, later renamed Oportunidades) was partly due to the presence of teachers and health workers who were temporarily under-employed from the financial-crisis-induced recession. However, I would not immediately assume that the government needs to provide these services. For, as we know in India (and in Africa), publicly provided primary education and health services are notoriously poor. The government needs to ensure that the services are provided, but the actual provision could be by the private sector (including the NGO and faith-based sectors). As you say, the government still needs to regulate the system, to ensure quality and standards. But I would not assume that regulation is more difficult than provision. Or even that the same government that cannot provide cannot regulate. Regulation requires different skills, and can also be subject to political capture, but it is less intensive in monitoring by the central authority (you don't have to monitor whether the teacher shows up every day, just whether the students pass an exam at the end of the year), and technology makes it possible for contracts to be transparent and therefore monitored by the public. The basic principle that cash transfers and vouchers enable poor people to hold providers more accountable remains. If you see resistance to the shift from public provision to cash-like instruments coming from teachers unions, for instance, it's a sign that the shift will end up making them more accountable to their clients, i.e. the poor. Best regards, Shanta

Add new comment