To cut waste and graft, collect taxes


Published 29.09.17

When local governments in Brazil improve tax collection rather than rely on grants to raise revenues, local officials help citizens, not themselves

Can we trust governments to spend revenues they receive in ways that improve the welfare of their citizens? The vast body of research on the topic from developing countries is pessimistic. When government revenues rise, typically there is little impact on public health, education or social infrastructure. Money is often wasted or diverted (e.g. Reinikka and Svensson 2005, Olken 2007).

But a common aspect of these studies is that they tend to consider only increases in non-tax revenues, such as grants from higher levels of government, earnings from natural resources, or official development aid. This is perhaps because it is hard to find variations in tax revenues that are unrelated to other determinants of public spending.

Governments might, however, spend tax revenues and non-tax revenues in different ways. Given the growing focus on revenue mobilisation in development, understanding whether an increase in a country’s capacity to tax means better public expenditure outcomes can help determine whether, and when, it is worth putting more effort into collecting that tax.

In recent research, I focus on local governments in Brazil to consider whether governments spend tax revenues better than non-tax revenues (Gadenne 2017). In Brazil, local governments control one fifth of public revenues. Their main expenditure is education, an area in which Brazil's performance is disappointing compared to countries at similar levels of development (Ferraz et al. 2012). Municipalities are in charge of primary education, and so they must take much of the blame for this. More needs to be done, and they could do more. there is also mounting evidence that Brazil’s local governments do not use increases in their non-tax revenues – whether grants from the federal government, or oil royalties – to improve local infrastructure. Instead, the extra cash is generally wasted or diverted (Caselli and Michaels 2013, Ferraz and Monteiro 2010, Brollo et al. 2013). In this context, it is worth asking whether tax revenues would be wasted or diverted in the same way.

Improving tax collection

I study a programme that helps municipalities increase their tax revenues. Local governments are in charge of collecting and setting the rates of two main taxes: a service tax, and an urban property tax. But most local administrations have little capacity to enforce tax payments. Municipal staff often rely on outdated tax registers, have little institutional memory. They assess liabilities by rule-of-thumb. The high costs of understanding and paying taxes, and the low penalties for tax dodgers, mean that many citizens simply don’t pay up. Some local officials have publicly admitted that there is a de facto permanent tax amnesty, and arrears will never be recovered (Afonso and Araujo 2006).

The Programa de Modernização da Administração Tributaria (PMAT) was launched in 1998 by the Brazilian Development Bank (BNDES) to improve this situation. It provides local governments with subsidised loans to invest in modernising their tax administration. There are three categories of expenditure for which the loans can be used:

  • Updating tax registers. Municipalities can improve their capacity to gather information on potential taxpayers. They can also invest in skills and software to analyse administrative data.
  • Stronger enforcement. This involves streamlining audit processes and recovering tax arrears.
  • Lower compliance costs. Introducing more ways to pay, more frequent, smaller payments, and simplifying taxpayers’ interactions with authorities.

The timing of the programme’s uptake allows me to disentangle a potential selection effect (municipalities that join the program may be the ones that already raise more taxes and spend their revenues better) from the causal effect of the programme on tax revenues. My results indicate that an investment of one Brazilian real in tax capacity led to an annual increase in tax revenues of roughly one real per year after five years. The extra revenue was invested in local public infrastructure. Using data on all municipal schools in Brazil, I find that the increase in tax revenues generated by PMAT led to an increase of between 5% and 6% in the quantity of municipal education infrastructure. There was also a significant improvement in an index of the quality of the infrastructure.

How does this compare to the impact of increases in non-tax revenues? I examine variations in non-tax revenues caused by a rule determining how large a federal transfer that municipalities were due to receive. Using this, I can estimate the effect of an increase in transfers of roughly the same amount as the increase in taxes generated by PMAT. The result: higher transfer revenues have no impact at all on local education infrastructure.

Better tax collection has other advantages too. Brollo et al. (2013) show that greater revenue transfers tend to result in higher corruption. In contrast, I find that a rise in tax revenues had no impact on graft.

Note that that the programme was voluntary, and only 300 local governments – those motivated enough to apply for it – took up the offer. So we can’t conclude from this evidence alone that tax and non-tax revenues would always be spent differently by all types of local government. Nevertheless, the results show that the project has increased the tax-collecting capacity of local governments that volunteered, and in turn has increased both the quantity and quality of local education infrastructure. Meanwhile, increases in federal grants to the average government have had no effect.

The right kind of decentralisation

The existence of a large ‘fiscal gap’ between local spending plans and tax revenues is pervasive across the developing world. On this evidence, efficient local tax collection is a necessary feature of successful decentralisation. Decreasing the fiscal gap would be one way that developing countries could improve the performance of local government.

The evidence also indicates that revenue mobilisation works in Brazil. This programme, in place for nearly 20 years, has provided long-term sources of funding for local governments. Beyond Brazil, these results suggest that development assistance to build tax capacity would create more public investment than grants or cash transfers.

Photo credit: Gilberto Marques/A2Fotografia.


Afonso, J R and E Araujo (2006), Local Government Organization and Finance: Brazil, World Bank.

Brollo, F, T Nannicini, R Perotti, and G Tabellini (2013), ‘The political resource curse’, American Economic Review, 103(5): 1759-1796.

Caselli, F and G Michaels (2013), ‘Do oil windfalls improve living standards? Evidence from Brazil,’ American Economic Journal: Applied Economics, 5(1): 208-238.

Ferraz, C and J Monteiro (2010), ‘Does oil make leaders unaccountable? Evidence from Brazil's offshore oil boom’, mimeo, PUC-Rio.

Ferraz, C, F Finan and D B Moreira (2012), ‘Corrupting learning: Evidence from missing federal education funds in Brazil’, Journal of Public Economics, 96(910): 712-726.

Gadenne, L (2017), ‘Tax me but spend wisely? Sources of public finance and government accountability’, American Economic Journal: Applied Economics, 9: 274-314.

Olken, B A (2007), ‘Monitoring corruption: Evidence from a field experiment in Indonesia’, Journal of Political Economy, 115: 200-249.

Reinikka, R and J Svensson (2005), ‘Fighting corruption to improve schooling: Evidence from a newspaper campaign in Uganda’, Journal of the European Economic Association, 3(2-3): 259-267.