Tax revenue collection as a share of GDP is only 15 to 20% in lower and middle income countries but over 30% in upper income countries. This gap is important, because it implies that developing countries have less tax revenue to spend on public goods such as infrastructure and good governance. Closing this gap is not as easy as it might superficially seem, because the gap is caused less by a choice of low tax rates and more by challenges associated with tax collection in developing countries: these include informality and misreporting, by both workers and firms (Besley & Persson, 2009). These challenges generate a vicious cycle, where developing countries remain poor because they are unable to mobilize revenue to invest in public goods. This highlights a key development challenge: what policies can build governments’ ability to raise revenue, and leverage that revenue to provide public goods, in developing countries?
The good news is that a growing literature has documented a number of policies that can increase tax revenue in developing countries. There are many excellent reviews of the broader literature — for instance: Pomeranz & Vila-Belda (2019) review experimental and quasi-experimental work with tax authorities, while Ulyssea (2020) discusses causes and consequences of informality among firms and workers. This blog non-exhaustively highlights some of the most recent empirical work in this space, stretching across interventions aimed at increasing the observability of the potential tax base (information, formalization) and directly increasing compliance (tax incentives, simplification and reminders, tax morale), with a focus on developing country contexts.
Observing the potential tax base
As households and firms may choose to underreport their income to evade taxes, independent reports of household and firm income are necessary in order to enforce taxes. For instance, Pomeranz (2015) uses randomized audits in Chile to show that as audited firms begin to report on their suppliers in value added tax filings, suppliers to audited firms pay more taxes. Enforcement can therefore cascade through the supply chain. Similarly, but targeting retail transactions in Brazil, Naritomi (2019) finds that offering lottery rewards to consumers for submitting receipts increased targeted firms’ reported sales by 21%. While these sorts of bottom up approaches can be effective, top down approaches may complement or substitute in other contexts. For example, Balan et al. (2020) find that local elites implementing tax collection in urban Democratic Republic of the Congo are able to leverage information about their communities to collect 43% more property taxes than provincial tax collectors. In complementary work in the same context, Bergeron et al. (2020) find that increased enforcement effort is complementary to increases in tax rates in raising tax revenue. However, top down approaches to enforcement may also backfire. In urban Haiti, Krause (2020) finds that, when provision of local public goods (in their context, garbage collection) is low, increases in enforcement may actually decrease property tax revenue collection.
Rather than underreporting their income, households and firms may remain informal and altogether evade reporting their economic activities to the government. Informality limits the ability of the government to maintain information trails that are necessary for tax enforcement. As reviewed in Bruhn & McKenzie (2014), there is growing experimental evidence that interventions can induce firms to formalize — in-person visits and training on formalization in Benin (Benhassine et al., 2018), inspections in Brazil (de Andrade et al., 2014), and incentives in Sri Lanka (de Mel et al., 2013). In some cases, the direct impacts of these interventions on government tax revenues were sufficient to cover costs, but only when the interventions were sufficiently well targeted.
When households and firms choose to evade, they may systematically misreport economic activities that are more difficult for the government to observe, implying that decisions about who and what to tax can have meaningful impacts on tax compliance. Bachas & Soto (2020) find that increases in corporate taxes on profits in Costa Rica cause firms to evade primarily by increasing reported costs, and that these responses are sufficient to decrease tax revenue, suggesting revenue taxes as an alternative. Unfortunately, Lobel et al. (2020) also find large evasion responses to revenue taxes in Honduras through decreased reported revenue, but these responses are much lower when the tax authority can independently observe firms’ revenue. In some cases however, evasion responses can actually increase optimal taxes — Bachas et al. (2020) find that because poorer households primarily shop in informal markets, consumption taxes in developing countries are mostly paid by richer households, offering an equity motive for relatively higher consumption taxes. Evasion responses are not limited to firms — Londono-Velez & Avila-Mehecha (2020) show that while wealthy Colombian individuals evade wealth taxes through offshore evasion, improved information on offshore accounts and amnesty programs for evaders can have large persistent impacts on tax revenue. However, not all noncompliance is driven by evasion motives – Brockmeyer et al. (2020) show that failure to pay taxes by property owners in Mexico is often caused by limited available liquidity rather than evasion.
Simplification and reminders
In many cases, tax evasion by households and firms may not be strategic, but instead economic agents may be unaware of when or how to pay the taxes they owe. Doing Business 2020 provides one example of the costs of this complexity, as the number of tax payments a medium sized firm must pay per year ranges from 3 to more than 60 across countries, and increased complexity is associated with increased perceptions of corruption. Consistent with this pattern, Cohen (2020) finds that that SMS reminder messages increased tax payment by 6% in Uganda, while De Neve et al. (2020) find that providing taxpayers with simplified information decreased delayed payments by 23% in Belgium.
Finally, households and firms may decide whether and how much to pay taxes as a function of how effectively they believe their tax payments will be used. Just as tax revenue enables government to invest in public goods, firms and individuals may choose to pay taxes only when governments spend wisely. In work mentioned earlier, Cohen (2020) shows that reminders are particularly effective at increasing tax revenue in Uganda when there have been recent government investments in services such as health and education, while Krause (2020) finds direct positive impacts of local garbage collection on tax revenue in urban Haiti. This behavior may in fact create positive incentives for government to invest in public goods — Gadenne (2017) finds that Brazilian municipalities spend tax revenue, but not grants, on school construction and maintenance. However and somewhat relatedly, direct impacts of public good investments on tax revenue through increases in the tax base are often sufficient to cover the costs of these investments, even ignoring tax morale effects — Hendren & Sprung-Keyser (2020) calculate this return across a broad range of policies in the United States.
Thanks for this interesting summary.
On the point around formalization and tax, you may want to also refer to EGAP's recent research effort studying interventions to increase formalization and tax outcomes. Overall, the outcomes are not as promising and suggest that formalization decisions are rational choices that weigh costs and benefits.