Electing educated leaders during democratisation: Evidence from Indonesia


Published 08.11.21
Photo credit:
Axel Drainville/flickr

The economic success of democratisation crucially depends on the education level of the newly elected local leaders

Following decades of debate, recent academic work shows that democratisation does cause an increase in national economic growth (Acemoglu et al. 2019). However, with the exception of a country’s level of development (Rodrik and Wacziarg 2005, Aghion et al. 2008, Acemoglu et al. 2019), very little is known about the conditions that determine whether and to what extent a transition to democracy is a success – Including the role of leaders’ characteristics. For instance, does the education level of newly elected democratic leaders matter, and if so, why? While Besley et al. (2011) find that the education level of national leaders matters for growth, they do not study political system changes and leave open the question of what highly educated leaders do better or differently. For example, educated leaders may implement more growth enhancing policies, although less educated leaders might focus more on policies that reduce inequality. 

In our recent paper (Pelzl and Poelhekke 2021), we contribute to this discussion by showing that the effect of democratisation on economic performance crucially depends on the education level of the newly elected local leaders, where several policy mechanisms explain this finding. We show that leaders’ education matters especially during transitions to democracy, and hypothesise that this finding may generalise to challenging times of political or institutional change.

Indonesia’s staggered transition to democracy

Our analysis focuses on Indonesia, which became the world's third largest democracy after the fall of President Suharto in 1998 following more than 30 years of autocratic rule. A unique feature of Indonesia's transition to democracy is that districts democratised in different years from 1999 to 2003. This was done in order to enable district mayors appointed by Suharto to finish their five-year terms before a new mayor was democratically elected. Since the end of term of the last ‘autocratic’ mayor is unrelated to district characteristics, this setting allows us to gauge the causal effect of a transition to democracy.

Manufacturing performance declines under non-college educated mayors

In terms of economic outcomes, we focus on the manufacturing sector, which has represented more than 20% of Indonesian GDP. As in many other developing countries, manufacturing has been targeted as the principal growth engine by the national government. 

Our analysis finds that manufacturing performance after democratisation is worse in Indonesian districts where the democratic leader does not have a college degree, irrespective of the last Suharto mayor's education level. 

Under non-college educated mayors, employment drops by an average of 5%, output by 15%, and total factor productivity by 1% in the first few years after the democratic election. In contrast, manufacturing performance remains constant under college-educated democratic mayors. We estimate these results based on panel data on the universe of Indonesian manufacturing plants with 20 or more employees, and by comparing districts that democratised in a given year to districts that did not yet democratise.  Figure 1 summarises the results and shows that the effects are immediate and increase in magnitude over time. 

Figure 1

Our results are robust to the inclusion of a wide variety of mayoral and district-specific control variables. This rules out, for example, that low leader education is simply a result of democratisation paired with low educational attainment of the local population. We also show that before democratisation manufacturing plants performed similarly across districts that would later elect a college graduate mayor and those that did not.

Alternative investments, such as in social programmes, are not higher

Despite national level recognition that manufacturing is a key sector, it is possible that democratic mayors without a college degree have been elected based on a different agenda and may prioritise other sectors at the local level. However, we do not find evidence for this: they do not spend more on items such as family welfare, health, housing, environment, religion, or education. Therefore, the poor performance of manufacturing in certain regions of Indonesia after democratisation does not only compromise the nation’s ambitious growth plans, but also shows more generally that democracy does not promote growth under all circumstances. 

What is it that non-college-educated mayors do worse?

While every policy debate benefits from evidence on the mechanisms that drive a certain relationship, such evidence has been surprisingly scarce around leader education and growth — both in a democracy and a more general context. We highlight multiple channels that explain the drop in manufacturing performance. 

Relative to college-educated mayors, non-college educated mayors: 

  • Raise taxation more which constrains business. The roots of higher taxation across all mayoral education levels can be attributed to the design of democratisation in Indonesia, which has by no means been unique from several other developing nations (Kulipossa 2004). Here, democratisation was accompanied by decentralisation, which was implemented nationwide in 2001 but assigned only minimal direct fiscal power to local governments. The qualitative literature identifies this decision, combined with poor governance across large parts of the country, as the primary reason for a proliferation of effectively unregulated fees and levies at the local level and the resulting constraints to doing business. Our finding that taxation rises more under mayors without a college degree suggests that the magnitude of such teething problems during times of political change depends on leader’s education. These observations and our matching plant-level results suggest important lessons for other countries that consider decentralisation and local elections, especially those where governance is relatively weak.
  • Provide less physical infrastructure. We find in a district-level survey that local physical infrastructure deteriorates, but only when the democratic mayor has no college degree. This holds for infrastructure availability and quality separately, and elements that may deteriorate or improve relatively fast such as “quality of telephone service” are affected more. 
  • Are more involved in corruption. Using hand-collected novel data we show that democratic mayors without a college degree are more likely to be involved in corruption cases.  

Our results on mechanisms confirm the finding of Keefer (2007) that young democracies engage in excessive rent seeking, under-provide public goods and are more corrupt relative to older democracies. We also highlight (insufficient) education as a driver of these issues.

Leader education matters most during democratic transitions

We find that the education level of the second democratic mayor does not have a statistically significant impact on local manufacturing, no matter the education level of the first democratic mayor. Moreover, the negative employment effects under non-college educated democratic mayors also hold when the last Suharto mayor has no college degree, such that the local mayor education level is unchanged. Leader education therefore appears most crucial as a country democratises, which is a novel result in the literature.

A choice for voters

In the first few years after democratisation, more educated leaders enact policies that yield better economic outcomes at a lower cost, compared to less educated leaders. Perhaps college graduates also better understand the detrimental impact of certain fees and levies on local business performance. 

Education is known to positively correlate with skills (Barro and Lee 2013). In the context of Indonesia, this relates to a 2003 statement where the Minister for Administrative and Bureaucratic Reform described the majority of civil servants as “undereducated,” where less than half “know what they are doing and do their jobs properly” (Webber 2006).

This quote and our results are drawn from a period of both democratisation and fiscal decentralisation, and mayor education matters less during subsequent democratic elections. Therefore, while Indonesia requires candidate mayors to have completed at least junior high school, our results do not necessarily suggest that a higher bar should be established.  

However, noting the better economic performance of more educated mayors during democratisation, voters may do well to select highly educated leaders in such times—and perhaps more generally during challenging periods of political and institutional change.


Acemoglu D, S Naidu P Restrepo and J A Robinson (2019), “Democracy does cause growth", Journal of Political Economy 127(1): 47-100.

Aghion P, A Alesina and F Trebbi (2008), "Democracy, technology, and growth", in E Helpman (ed.), Institutions and Economic Performance, Harvard University Press.

Barro, R and J-W Lee (2013), “A New Data Set of Educational Attainment in the World, 1950-2010”, Journal of Development Economics 104: 184-198.

Besley T, J G Montalvo and M Reynal-Querol (2011), “Do educated leaders matter?”, The Economic Journal 121(554): 205-227.

Keefer P (2007), “Clientelism, credibility, and the policy choices of young democracies”, American Journal of Political Science 51(4): 804-821.

Kulipossa P F (2004), “Decentralisation and democracy in developing countries: an overview”, Development in Practice 14(6): 768-779.

Pelzl P and S Poelhekke (2021), “Electing Educated Leaders during Democratization: Evidence from Indonesia”, CEPR Discussion Paper 16486.

Rodrik D and R Wacziarg (2005), “Do democratic transitions produce bad economic outcomes?”, American Economic Review 95(2): 50-55.

Webber D (2006), “A consolidated patrimonial democracy? Democratization in post-Suharto Indonesia”, Democratization 13(3): 396-420.