One of the strongest arguments for investing in the tax capacity of developing countries is the supposed participation dividend of taxation. Beyond providing more tax revenue with which the government can supply essential public services, tax collection is thought to stimulate citizen engagement in the government, and ultimately, political accountability. The accumulated evidence of a participation dividend of taxation in developing countries is, by now, considerable, spanning diverse settings and research methods. Building on past evidence from cross-country regressions, in-depth case studies, survey experiments, and lab-in-the-field experiments, a new paper, published in July in the Quarterly Journal of Economics, provides further evidence from a randomised policy experiment in the Democratic Republic of Congo.
Evidence from the 2016 Property Tax Campaign in Kananga
In 2016, the Provincial Government of Kasai Central sought to raise revenues by extending the property tax net well beyond the small set of firms downtown that had paid the tax in the past. It organised a citywide door-to-door tax campaign in which collectors registered and solicited the property tax from citizens.
Crucially, the government randomised the roll-out of this campaign on the neighbourhood level. Thus, a little over half of Kananga’s 431 neighbourhoods received tax collectors in 2016, while the rest were slated to receive them in the future. The government chose to randomise which neighbourhoods were in the first phase of the campaign in order to evaluate its impact rigorously. Comparing “treatment” neighbourhoods, which received collectors, to “control” neighbourhoods, which did not, reveals the causal effect of the tax campaign on tax revenues and on political participation.
According to the government’s property tax database, the campaign increased property tax compliance by over 11 percentage points. Although the majority of citizens still avoided paying the tax, property tax compliance reached levels on par with national capitals in more affluent African countries. It was an impressive achievement for the city’s first large-scale property tax campaign, and the government anticipates further increases in compliance in the future.
The tax campaign also catalysed citizen engagement with the provincial government. Citizens living in treated neighbourhoods were more likely than those in control neighbourhoods to attend town hall meetings hosted by the provincial finance ministry. They were also more likely to submit anonymous evaluations of the government to a locked drop box in downtown Kananga. Citizens spent on average a full day’s earnings to participate in these ways, demonstrating the value they placed on having a voice in the provincial government.
Importantly, citizens appear to have chosen to participate in order to bargain with the provincial government for a better fiscal deal. “Why should the inhabitants of Lukonga pay taxes,” one participant asked, “when the roads are in such disastrous condition?” They also demanded more transparency and accountability over spending. “The provincial government should do more,” noted one individual, “and inform us how this money will be spent on public infrastructure and not wasted on other things.”
The paper thus provides evidence from a randomised policy experiment in a resource-poor country with low state capacity that tax collection increases citizen participation and demand for accountable governance. The results support the idea that investments in tax capacity can both enable greater future public goods provision and also help usher in more inclusive governance.
Paths of Future Inquiry
There are of course important limitations to the external validity of these findings and many related questions that would benefit from further research.
First, the 2016 collection campaign in Kananga was the first time in which many citizens had been solicited to pay the property tax. It reflects a discrete change in the implied social contract between citizens and the state. It cannot teach us about how further tax collection efforts will affect participation. Does the participation dividend of taxation erode as citizens become accustomed to tax appeals, or are subject to marginal increases in enforcement? An ongoing study of a multi-faceted property tax reform in Sierra Leone tackles these questions, but there is ample space for more research in this area.
Second, marginalised groups can face heavy tax burdens yet are often denied a seat at the bargaining table. In Kananga, 75% of citizens who attended town hall meetings were men. Further evidence on the heterogeneous effects of taxation on participation by gender, ethnicity, and socioeconomic status is greatly needed.
Third, the property tax is a direct tax, and the collection campaign in Kananga was highly visible to citizens. Indirect taxes are less visible, manifesting in higher prices paid by citizens for goods and services they consume. Do indirect taxes induce a weaker participation dividend? A recent working paper examines country-level regressions and lab-in-the-field experiments to suggest that, indeed, indirect taxes lead to fewer citizen demands because they are less visible. Researchers could examine this issue further by exploiting idiosyncrasies in the roll-out of value-added taxes over the past few decades, or more generally by studying how variation in the salience of government tax enforcement shapes the political response among citizens.
Fourth, the 2016 campaign was implemented by the provincial government in its provincial capital. The government that received and spent the new tax revenues was geographically close to those who were asked to pay taxes, and the flow of tax funds was straightforward. By contrast, many taxes, including the income tax, are collected locally and remitted to the national level before (in theory) returning to the provinces in the form of retrocessions. In such cases, it is less obvious whom citizens should hold accountable. How the tax-spending link impacts the participation dividend is particularly important given the fact that income taxes tend to account for an increasing share of revenues as countries develop.
Fifth, firms are key taxpayers in developing countries and often have a powerful voice in politics. How increased enforcement of firm taxation shapes their political engagement is an important issue in the tax-accountability literature that has received less attention. “Bringing small businesses into the tax net can help secure their participation in the political process,” wrote the IMF in 2011, “and improve government accountability.” Further evidence testing this claim would be valuable, especially in light of recent studies highlighting significant barriers to engaging small businesses in practice.
Sixth, in sub-Saharan Africa and other developing countries, a range of different actors are involved in tax collection. Customary authorities, for instance, have important roles in formal and “informal taxation” in the DRC and many other countries. How the engagement of such actors shapes the participation dividend of tax collection is another topic on which further evidence is needed.
Finally, a major topic is how citizens respond after they observe how the government spends newly collected tax revenues. An ongoing study in Pakistan, for instance, examines responses to (a) any public goods provision, and (b) the ability of citizens to choose what public goods their tax dollars will be spent on. Similar research around new taxes and participatory budgeting is being carried out in Sierra Leone. These ambitious studies will hopefully spark a new empirical literature exploring how tax-based social compacts emerge and evolve over time.