By Dr Ross Harvey, director of research and programmes at Good Governance Africa (GGA)
A view that appears to be gaining traction in some policy circles is that African countries need a benevolent dictator. The argument runs along the lines of suggesting that we’re not quite ready for democracy. Due to democracy’s apparent inability to strengthen institutions and produce economic growth, we should resort to a kind of autocracy that delivers material dividends. Some have framed this as a choice between inclusive or effective governance. But even this framing presents a false dichotomy. Effectiveness is a function of inclusiveness, though admittedly under the right conditions of a relatively cohesive state.
These debates are not merely theoretical. Empirical data from the Afrobarometer 2023 survey, for instance, showed that 72% of South African respondents said they were willing to forego elections if it meant increased security and material wellbeing. This correlates closely with 63% of respondents to the same survey indicating that they did not feel close to any political party. From 2000 to 2015, respondents felt a strong sense of ‘partisanship’ (feeling represented by a party), but the post-2015 drop-off to 37% correlated with a sharp increase in those willing to forego elections. The drop in partisanship seemed to precipitate the willingness to abandon democracy. Until 2015, the trends largely moved together. Thereafter, they diverged radically.
There are all kinds of reasons for the steep drop in partisanship; state capture was exposed but prosecutions have been slow. Simultaneously, the institutional fabric required to attract investment that could generate labour-absorbing growth was being badly frayed. This is the very fabric required to ensure swift prosecutions. It hasn’t helped that 2022 and 2023 were the worst load-shedding years since 2008. Voter turnout plummeted in line with the Afrobarometer survey results, and in the most recent elections was a poor 58%. So, do we need to drop all this idealistic democratic flag-waving and embrace aspirant autocrats who will set everything to right with an iron fist?
Many point to Rwanda as a model example. Not those who’ve fled for exercising judicial courage, of course, but armchair observers who see apparent value in the empirical results. Everyone knows that the elections are a sham, and one could thumb suck a winning number for Kagame (though the minimum floor is 80%).
But let’s examine the results against Kenya and Zambia, for instance, two democracies in Africa that have withstood the test of power turnovers. I could have chosen any number of metrics, but there are at least three quite important determinants of whether a country has long-run success probability or not. First is access to electricity for the rural population, as this indicates ability of the state to deliver public services outside of urban areas that can typically raise their own taxes, and electricity is a necessary condition for economic growth. As the graph below indicates, Kenya outstrips Rwanda by a country mile. By 2021, 68% of the Kenyan rural population had access to electricity, whereas Rwanda only reached 38%. Zambia had been beating Rwanda until 2013, after which it fell behind and is yet to catch up.
Second, Rwanda’s unemployment levels have climbed from a steady rate of 12% in 2017 to 15% in 2023. Kenya’s levels also rose during that time, but from 2.76% to 5.6%.
And Zambia’s rate, for all the country’s volatility and poor leadership until Hakainde Hichilema entered office (in 2021), plummeted from a high of 16% in 2005 to beat even Kenya by 2023 with a low of 4.22%. An improved investment environment for mining will help to bring that figure even lower.
Thirdly, the “Country Policy and Institutional Assessment” (CPIA) “public sector management and institutions cluster average” scores (1=low and 6=high) of each country reveal that this is the only arena in which Rwanda wins, which is not nothing. The cluster “includes property rights and rule-based governance, quality of budgetary and financial management, efficiency of revenue mobilisation, quality of public administration, and transparency, accountability and corruption in the public sector.” This is more important for us to understand than ‘outcome’ data such as economic growth, given how pivotal a country’s institutional framework is for catalysing and sustaining growth and poverty reduction.
Rwanda’s score has climbed from 3.3 since 2005 and stabilised at 3.8 (by 2022). Kenya started in the same place but has only moved up to 3.5 (by 2022). Zambia, by contrast, has declined from 3.2 in 2005 to 2.9 by 2022. The hope is that under Hichilema’s leadership, the country will implement the right institutional changes and reverse the decline.
By the above, one can understand why the Rwanda model seems attractive. When one considers the long-term risks, though, it is harder to understand. Kagame is a ruthless operator, and civil liberties are deeply repressed in Rwanda. Aspirant autocrats – especially when they become entrenched dictators – start to make mistakes that hurt the citizens who were perhaps willing to forego liberties such as a free press, an independent judiciary and a credible vote. “Big men” surround themselves with yes-men and avoid any checks and balances. Accountability goes out the window. Even patronage beneficiaries start to live in fear.
So, what to do then, given that democracy appears to not always deliver? An under-appreciated 2006 paper by Michael Ross - “Is Democracy Good for the Poor?” – is worth consulting: “Over three decades [1970-2000, in which there was a radical reduction in child mortality rates across the world], there was also a dramatic rise in the prevalence of democracy; yet we find little evidence that the rise of democracy contributed to the fall in infant and child mortality rates. Democracy unquestionably produces noneconomic benefits for people in poverty, endowing them with political rights and liberties. But for those in the bottom [income] quintiles, these political rights produced few if any improvements in their material well-being. This troubling finding contradicts the claims made by a generation of scholars.”
However, a 2019 paper by a group of world-class economists (Daron Acemoglu and James Robinson among them) ran several complex econometric regression models to try and isolate the impact of democratisation on economic growth. They found that “democratisations increase GDP per capita about 20% in the long run.” This is mostly likely due to democracies (on average) “enacting economic reforms, improving fiscal capacity and the provision of schooling and healthcare, and perhaps also by inducing greater investment and lower social unrest.” Clearly, as Michael Ross points out, under-5 mortality (a sensitive indicator of poverty) has not necessarily reduced due to democracy being better than autocracy at reducing poverty per se. Nonetheless, the results from Acemoglu and his co-authors show that the economic benefits of democratisation are not heterogeneous at different income levels. In other words, countries with more inequality and lower average income levels did not fare worse than their richer or more equal counterparts in terms of long-run growth.
The idea that democracy is bad for growth at early stages of economic development has been debunked. Dictators can do great damage, as Kagame continues to do, even while having made some incremental gains on the CPIA score. Civil liberties, to the contrary, can equip citizens to call for accountability that will likely produce pro-poor growth (though it’s never a guarantee).