My anchor in this topic is Tim Besley’s ongoing work on tax capacity as an investment made only if it is in the interest of a country’s “selectorate”: the small network of people who wield power of veto (see, e.g., Tim Besley and Masayuki Kudamatsu, “Making Autocracy Work,” LSE STICERD Research Paper, 2007). Lucy Martin switches the focus as to which state capacity matters from the ability to tax to the ability to deliver the services that would justify it. Her thesis is that in the conditions of low state capacity such as prevail in much of Africa, democratization weakens the incentive of governments to tax, a proposition formalized into a model and tested against the modest data available. The kernel of the argument is that governments that lack the capacity to deliver the public services that could justify higher taxation are wary of taxing their citizens. That wariness arises because the voting power of citizens in a democracy threatens government survival.

However, a causal connection from the lack of some state capacity to increased reluctance of a democratic government to tax could arise from different deficiencies in state capacities. Consider the introduction of limited democratic rights into a previously autocratic system. Suppose that the rights are less than implied by the electoral honesty required to threaten an incumbent ruler with defeat. Nevertheless, they inhibit the capacity to oppress potential opponents. Herodotus tells the story of a young Greek dictator seeking advice from an experienced one: the advice is to cut down tall poppies before they become a threat. Such Stalinist pre-emptive precautionary state violence is curtailed by even a token degree of democracy. That was often the modest form of democracy that African rulers conceded under pressure from donors empowered by the collapse of the Soviet Union in 1990. In the language of economics, such quasi-democracy was “technical regress” in coercion.

Martin relies quite heavily on Uganda for her evidence, where Yoweri Museveni is now in his 40th year as president. His repeated electoral triumphs may be gratitude for the undoubted state capacities he has built in both taxes and services. However, tax is only up from an unviable 6% to a modest 12%, so he is clearly wary of voter disapproval. Perhaps this shows a democratic ruler successfully balancing tax resistance with affordable service delivery. But if so, he is unique: no other leader of a fully democratic country has ruled that long. More likely, like many other countries, Uganda is partially democratic: the opposition is permitted to contest an election but not to win it. Why might this curtail taxation despite there not being a threat to regime survival? I suggest that the reduction in repression has exposed the state to the repeated embarrassment of vocal opponents who would pounce on increases in taxes regardless of whether they had merit. Like voters elsewhere, they are reluctant to link the services they crave to the taxes needed to pay for them. This can be overcome, as I will show, but partial democracies make it difficult because the opposition never faces the responsibility of power.

Not only is the distinction between autocracy and democracy a continuum, so is sovereignty. Strategic Taxation correctly describes the radical shift in the composition of African taxation from trade to VAT but omits that it was driven by IMF conditionality. Recognizing it enriches Martin’s distinction between weak and strong state capacity. The IMF favored VAT because in high-capacity OECD states it is less “distortionary” than sales taxes. But its visiting teams often lacked the detailed knowledge of how VAT might operate in a low-capacity African state. In an OECD economy, the value of a product gradually accumulates as it passes from one firm to another. VAT avoids taxes at subsequent stages of production being levied on the tax-inclusive price. It does so by rebates to firms in return for evidence of their tax payments on their purchases. In the rudimentary economies of low-income Africa, the compounding of taxation is a less severe problem, but in the context of weak administrative capacity, the honest conduct of a rebate system is difficult. In my estimates, in six such countries the introduction of VAT reduced net revenues as the rebates opened opportunities for scams.

As this illustrates, African governments have had only limited agency. Between the fall of the Soviet Union and the rise of China, the USA and the International Financial Agencies over which it had considerable influence had excessive power over them. The IMF adopted an explicit approach, “buying” unpopular policy reforms in return for cheap loans, while playing the scapegoat for voters to blame. This lowered the political cost of raising taxes, which would perhaps be helpful within Martin’s framing of the problem. If so, I would disagree: if the reforms are unwise, the government should reject the money; if they are good (as they often are), the government should explain the case to citizens, building a more policy-literate population. Denmark is, by many criteria, the world’s most successful state. Its Social Democrat Premier has repeatedly been re-elected despite taxation at 53% of national income. Denmark’s premiers have long invested in a financially literate electorate. Now that Danish defense expenditure needs to jump, it is being financed by painful reductions in social spending, justified to voters by “you can’t spend the same Kroner twice.”

The world’s least Danish society was Zaire under Mobutu Sese Seko, a coup leader in power for 30 years. He faced zero democratic pressure yet imposed negligible taxation and provided no services or infrastructure. This challenges Martin’s model because it was a coherent response to his situation. His zero-infrastructure strategy made him safer: a rebel leader would need roads and airfields to mount a coup. He built no civil service capacity because it would constitute an educated class opposed to his plunder of the country’s resources. Instead, he relied on his illiterate soldiers. Their job was to stop highly visible goods like food and drink from entering the ports unless the importer could show a crested letter with the presidential seal, which he sold to enrich himself.

This is an illustration of the larger issue that context is so important for the analysis of African taxation that the simplifications inherent in a grand formal model may be premature until we have a richer stock of fieldwork. The transition away from near-zero taxation, such as currently faced in the Democratic Republic of the Congo, may have very different solutions to that faced by Uganda in moving from 12% to 25%. Both may be entirely different from getting to the 45% of European welfare states. One handy concept introduced by Cecelia Heyes is a ‘cognitive gadget’: a non-obvious technique which once a society has hit on it, is retained and from which similar societies could learn (Cecelia Heyes, Cognitive Gadgets: The Cultural Evolution of Thinking, 2018). Pakistan pioneered the payment of incentive bonuses to tax collectors; DRC pioneered using local chiefs to increase tax collection. Neither is learnable from Denmark.

Another handy concept is Margaret Levi’s “community-of-fate” (see Margaret Levi, Expanding the Community of Fate by Expanding the Community of Care, Daedalus, 154(1), 2025). That explains Denmark’s high taxes: willing compliance by the many in the objectives of the state has enabled effective enforcement on the recalcitrant. Africa’s history of colonialism destroyed such communities, usually bequeathing the office of head of state without either the coercive power or the authority of tradition needed to enforce taxation. Africa’s leaders have inherited a challenge for which they have needed more evidence and deeper analysis: this book is a valuable contribution to both and a foundation on which further research can build.