Shadow economies present a major problem for governments worldwide, particularly in developing countries. Economic activity that takes place in shadow economies is usually untaxed and unregulated, reducing governments' fiscal revenues and eroding state power.
A discussion paper disseminated by the Institute for the Study of Labour (IZA) in Bonn in December was one of the most comprehensive assessments thus far of the size of shadow economies worldwide. It covered 145 countries, using data up to 2002-03.
The report found marked differences in the size of shadow economies in different groups of countries:
OECD
In the 21 OECD countries researched, the (unweighted) average size of the shadow economy in 1999-2000 was 16.8% of GDP. This decreased to 16.3% in 2002-03. Greece, Italy and Spain had the largest shadow economies (as proportions of GDP) and the United States, Switzerland and Japan the smallest.
Asia (excluding Japan)
The average size of the shadow economy for the 27 Asian countries covered increased from 28.9% of GDP in 1999-2000 to 30.8% of GDP in 2002-03. Thailand and Cambodia had the largest shadow economies in Asia in 2002-03, while Hong Kong, Saudi Arabia and Singapore had the smallest.
Among developing countries and the 'transition' countries of Central and Eastern Europe, the average size of shadow economies was over twice the OECD average and rising:
Africa
For 37 African countries researched, the average size of the shadow economy rose from 41.3% of GDP in 1999-2000 to 43.2% in 2002-03. At the latter date, Zimbabwe, Tanzania and Nigeria had the largest shadow economies (as a
proportion of GDP), while South Africa recorded the smallest shadow economy at both points in time.
Latin America
The average size of the shadow economy of the 21 Latin American countries covered rose from 41.1% in 1999-2000 to 43.4% in 2002-03. The largest of these were in Bolivia, Panama and Peru, with Chile and Costa Rica having the smallest.
Central and Eastern Europe
For 25 Central and Eastern European countries, the average size of the shadow economy rose from 38.1% of GDP in 1999-2000 to 40.1% of GDP in 2002-03. Georgia and Azerbaijan had the largest shadow economies, and the Czech Republic and Slovakia the smallest .
Definition concerns
A general pitfall in estimating the size of the shadow economy is that there is no agreed definition of what exactly to include. The IZA estimates are based on a narrower definition than many studies. They include all 'market-based legal production of goods and services' that are intentionally hidden from the state in order to avoid:
-taxes;
-social security contributions;
-compliance with labour market regulations; or
-administrative procedures.
This definition does not include criminal activities such as theft or the drug trade, or output within households.
Estimation techniques
Many approaches have been used to estimate the size of shadow economies:
Declared vs actual income. Survey based comparisons are undertaken of declared and actual income (which includes income not declared for tax reasons). However, samples for such 'fiscal audits' are usually biased towards the group of people
among whom tax fraud is suspected. Moreover, obtaining accurate estimates of individuals' actual incomes can be extremely difficult.
National expenditure vs national income
The discrepancy between national expenditure and national income could serve as a measure of the shadow economy. However, expenditure (as well as income) is often unreported, so the discrepancy understates the size of the shadow economy. Demand for currency. Activities in the shadow economy presumably involve mainly cash transactions. Therefore, shifts in the demand for currency can be used to proxy the shadow economy. However, such shifts may also occur for other reasons,
undermining the usefulness of this proxy.
Combined approach
The IZA paper used a more complex (though not new) method: an 'unobserved variable' estimation model combined with previously existing data obtained using the 'currency demand' estimation technique. The 'unobserved variable' model treats the shadow economy as an 'unobserved variable', which has several causes (such as tax rates) and indicators (such as official employment rates). The attractiveness of this model lies in the large number of causal and indicator variables incorporated. Ironically, this is also its main weakness, as the explanatory power of each causal and indicator variable of the shadow economy is itself uncertain.
Policy implications
The approach used in the IZA study illustrates the increasing sophistication of shadow economy estimates. While these remain open to criticism, there are two major implications of the existence of shadow economies for governments of both developed and developing countries:
Multiple equilibria. The shadow economy clearly reduces the tax take and social security contributions to the state. Given fiscal needs, this can lead to an increase in tax and social security rates, thereby further increasing the incentive to join the shadow economy. Governments need to be aware of the risk of becoming trapped in this 'low equilibrium', rather than a 'high equilibrium' that includes lower tax rates and a smaller shadow economy.
Regulatory enforcement
Studies have shown that higher levels of regulation are correlated with a larger shadow economy. However, improving the implementation of regulations would probably reduce the size of the shadow economy. Policy should thus emphasise enforcement of existing regulations rather than the adoption of new regulations.
Estimations of the size of shadow economies are increasing in sophistication, but remain highly uncertain. Nevertheless, it is clear that the average size of the shadow economy is substantially lower as a proportion of GDP in developed than in developing countries. However, even in developed countries, shadow economies seem to be sufficiently large to warrant increased attention from policymakers.
By: N. Peter Kramer
