Our Programmes



Sign up to our newsletter.

Follow ODI

Financing the State: Tax and Revenue Issues

Time (GMT +01) 16:00 17:30


Alice Sindzingre, Research Fellow, National Centre for Scientific Research (CNRS, Paris)/Research Associate,
School of Oriental and African Studies (SOAS)


Jonathan Di John, London School of Economics (LSE)

In her presentation Alice Sindzingre made two main arguments regarding the role of taxation in creating a developmental state:

  1. Taxation systems are in themselves insufficient to explain the creation of a developmental state. In her view what is important is a state that credibly commits and intervenes through policies that are directed towards growth, not ones that recycle the national wealth through taxation. Such a state addresses coordination failures, reallocates factors of production and creates coalitions between the state, private firms and civil society.

  2. Effective policies and institutions need to be credible. Society at large must perceive commitment on the part of institutions for them to work efficiently. In her view, such perceptions of credibility and commitment are created through endogenous processes. External donors are ill equipped to facilitate these processes.

Sindzingre substantiated these arguments in her presentation by discussing 4 aspects of developmental states:

  1. Concept of the developmental state: Sindzingre stated that the North and East Asian (EA) states did not rely on high levels of tax collection for growth. Instead they focussed their efforts on active development strategies and industrial policies such as targeted taxation, limitation of foreign shareholdings, incentives for the banking sector, firm financing and training in technology. In the East Asian case the tax to GDP ratio was an irrelevant proxy of state intervention. The state relied on lessons from the 'founding fathers' of development of intervening to reallocate factors of production in a manner oriented towards development and creating conditions for co-ordination between sectors and economic agents to facilitate learning processes. In successful Asian states political rent seeking was tuned to market sanctions. In sub-saharan Africa (SSA), therefore, the constraints to development are not "wrong relative prices" but policies that are not conditional on developmental objectives. To create a developmental state, ingredients of development cannot be considered in isolation but only in combination with economic, political and social factors.

  2. Features of taxation in low income countries: the case of Sub-Saharan Africa: The speaker then focussed on particular features of taxation systems in SSA which hinder developmental outcomes:

    a) incentives to collect taxes are eroded due to an over-reliance on external resources;
    b) dependence on primary commodities which suffer from price volatility. The "natural resource curse" can also prevent nation building, create incentives for corruption and is associated with higher inequality;
    c) High taxes on export crops (e.g. Ghana where average tax rate on cocoa was 58% in 1990s, compared with 24% in 1900s) accompanied with ineffective stabilisation boards;
    d) heavy reliance on trade taxes for government revenue which constitute between a quarter and a third of government revenues compared with 2% in high income countries;
    e) institutional, informational and political problems in shifting from "easy to collect" taxes to "hard to collect" taxes such as VAT as advocated by the IMF.

    The EA case demonstrates that the size of governments and the level of public spending do not influence development outcomes as much as the macroeconomic context. The public sector in SSA is not excessively large. Sindzingre listed some reasons explaining why public spending in SSA has not performed:

    a) increasing share of wage bill due to fiscal squeezes prescribed by International Financial Institutions (IFIs) at the expense of investment and maintenance spending
    b) funds increasingly directed to social sectors as advocated by IFIs while the link between social spending and growth remains controversial
    c) limited public spending on social security which can further create distrust of the state as individuals rely on "informal social protection". This contrasts with the EA case where the state did not divert high levels of resources to social protection rather it provided social welfare selectively to build political support.

  3. Consequences of reforms on formation of developmental states in SSA. The speaker touched on the following consequences of trade liberalisation in SSA:

    a) Tax revenues in SSA decreased from 16.3% of GDP in early 1990s to 15.9% in early 2000s. IMF studiesthemselves show that only 30% of losses in revenue were recouped from domestic indirect taxes;
    b) no reduction of vulnerability to external shocks;
    c) no diversification of exports from natural resources;
    d) fiscal squeeze reduced the capacity of public policies towards export diversification and infrastructure spending which are important for forming a developmental state;
    e) intensified international tax competition in a rush to attract FDI while EA states were cautious to FDI.

    Reforms also envisaged autonomous agencies for revenue collection but political economy considerations may prevent such an agency from being truly "autonomous" in Africa. Reforms also assumed that private entities would be more efficient whereas in reality this has not really happened. More generally IFIs ignored ingredients of state building of the development states and the "Scandinavian model" of "more state and more market" in its optimism over the minimal state.

  4. The effects of aid in terms of taxation and building the developmental state
    No amount of aid can create a development state. The speaker stressed that during the early stages of development what is important is not external financing but co-ordination devices and long term strategies that efficiently reallocate factors of production. The IMF (as opposed to the World Bank) is sceptical of the overall effectiveness of aid. Rajan and Subramanian have concluded that aid has a negative relationship with growth particularly in aid-dependent countries.

    Fiscal effects of aid may be negative because:

    a) tension between conditionality and ownership;
    b) aid provides incentives to reduce tax effort;
    c) aid dependence can weaken political institutions. If governments raise their revenues from aid they become more accountable to donors than citizens;
    d) aid fungibility;
    e) composition of aid: higher spending in social sectors can worsen the fiscal squeeze is aid is diverted towards projects that may not contribute to growth;
    f) aid volatility: the average volatility of aid is 40 and 20 times higher that that of revenue and falls short of pledges by more than 40% especially for poorest countries.

    The delivery of aid essentially involves multiple actors, and there is a lack of incentives for them to co-operate.

Jonathan di John approached the role of taxation in creating a developmental state from a historical and political economy perspective. He discussed the:

  1. Role of threat & conflict: in creating developmental states, which are now High Income Countries. The main literature for this idea comes from Charles Tilly who hypothesised that states initially collected taxes to build armies for survival but during that process also developed administrative capacity with which they were able to thrust into society more deeply. These efforts to finance war led to varying patterns of bargains between the state and interest groups particularly merchants, landlords and in some cases directly with the peasantry. Threat also ensured implementation of agrarian reforms such as peasant revolutions in the Scandinavian states, which ensured that agrarian reforms were more consolidated. Agrarian reform is important for state formation because most resource mobilisation is from states using surplus from agriculture to more industrial sectors; widened markets are created for industrial goods and political legitimacy is created.

    He posed the questions: what kinds of threats in the current world can lead to this type of developmental tax collection for e.g. globalisation? Are there some kinds of wars that are more developmental than others e.g. the ones in SSA have led to a breakdown rather than a construction of states?

  2. Resource mobilisation in Developing Countries: Developmental states mobilise resources through forced savings, agricultural marketing boards, state enterprises and these non-tax features are more if not most important in leading to a developmental state. In 1997-2000 there was not much difference in tax to GDP ratio between for instance between EA and South Asia however there were large differences between savings rate: saving as a % of GDP in EA between 1990-02 was 31.2 % compared with 16.7% in South Asia. There is a robust relationship between economic growth and public savings. The State in EA mobilised public and private savings through for instance restrictions on consumer credit or mandatory pension contributions.

  3. Composition of taxes: Jonathan di John produced data to show that within middle income countries there are wise variations in the composition of taxes when aggregate tax levels are the same for e.g. 1997-2002 Latin America collects 1% of taxes from income and property taxes as a percentage of GDP and East Asia it is 4 times as much. This is not due to differences in per capita income but represents the leverage state has over its elites.

  4. Inequality: With trade liberalisation IFIs recommend substituting the reduction in tax revenues from trade taxes with an increase in taxes such as VAT. This recommendation rests on the idea that VAT taxes are easier to collect. However in regions like Latin America where there is a lot of inequality introduction of VAT can be destabilising because it is essentially a very regressive type of tax.

  5. Politics and tax reform: IFIs recommend that revenue collection authorities are more effective when they operate autonomously from the state, as a commercial entity at arms length from the government. There is some evidence in Africa and Latin America that although autonomous revenue authorities (ARA) can be instrumental in initiating reform they may not be sustainable. This technical approach abstracts from politics in three ways:

    a) why such reforms were politically feasible in the first place is not addressed;
    b) there is little analysis of why such autonomy is acceptable to relevant political coalitions over time and
    c) there is no accepted definition of autonomy. Autonomy can never be complete when there are inter-dependencies among agencies.

In Uganda such reform was unsustainable due to political strategy of anti-party politics, which rendered the tax authority vulnerable to shifting policies and coalitions. This kind of tax reform has often taken place in the context of a leader who rises from the outside so is less likely to be hostage to power groups. Then when an unpopular tax is introduced then the leader can blame the tax agencies and so it becomes more a target of politicising than before.

Di John suggested two further reforms:

  1. Rethink the policy of exempting high-income expatriates from paying taxes despite the facet that they often earn 100 times the national average salary (e.g Afghanistan)

  2. Aid would do more if channelled through the central government otherwise it can create a dual public sector.

The discussion that followed the speakers' presentations brought up the following ideas:

Q: How can IFIs dictate very specific policies when we are still not generally sure what work in generating development outcomes?

A: Most of the conclusions of the IMF are based on cross country regressions which do not take into account political economy factors. External donors should at a minimum attempt not to do harm in a country environment. Most external interventions are based on a model which applies to rich countries where contain well financed, legal interest groups. LICs have very underdeveloped labour union, large subsistence economies, formal, legal mechanisms are not available and therefore external donors should seek to influence development patterns where formal mechanisms are the exception not the rule.

Q: Countries in Africa are richer because official statistics don't capture the informal economy that doesn't pay taxes. Is VAT the answer?

A: "Informal" economy is a misleading team when the majority of the economy lives according to those rules. Most studies show that informal economy pays taxes. In fact African economies do not have a dualistic economy but a continuum of activities some of which fall within the formal sector and some within the informal sector so VAT is difficult to implement. In order to create development it is not the type of tax that matters but the production strategies employed by the state. Tax literature is actually divorced from production strategies so VAT has little do to with creating development outcomes.

Q: There isn't a shortage of conflict in Africa so why hasn't Charles Tilly's hypothesis played out? Are there any other answers explaining development?

A: In Africa states are elites are not aligned to defend common borders. Instead there are regional wars which are about preserving neo-patrimonial networks rather than building states. Perhaps borders should be open for debate to provide incentives for leaders to provide functional governance.

Q: Is building state capacity part of the equation and if so where can we start to build it?

A: Best way to look at state capacity is to compartmentalise its functions for e.g South Africa collects more taxes than South Korea but South Korea has a better industrial policy. It is essential to think of policies that force leaders and elites to make bargains. If elites exit the system by not paying taxes and paying for private services there is no one to put pressure on governments and NGO's can fragment and marginalise the system further.

Q: What sort of conditions, threats and incentives might move states in Africa in the direction of developmental states?

A: Political economy, rules and incentive structures matter. The time frame of leaders can be the same but one may be politically motivated to perform while the other may be threatened by growth.

Q: If aid is unpredictable isn't it better for governments to build their own resources so there is greater incentive to tax not less?

A: Even with the harmonisation and alignment agenda donors are not playing inherently a co-operative game because it is not in their interests to pool resources..

Q: What can you advise international policy makers to do? Where does the initiative start? Who is going to create that if the groups are marginalised?

A: Current aid structures provide incentives for donors to continue as before. The reality is that in countries causes and motives are multiple. You cannot simply look at economics to explain the effectiveness of a structure. Structures of production are a result of political actions, social norms and history. With these multiple determinants no donor can predict how people will process reform. Any reform has to be designed on a case by case basis.


In this meeting from the '(Re)building Developmental States: From Theory to Practice' series, Alice Sindzingre discusses the role of taxation in creating a developmental states.