In this chapter, we begin by summarising the central contributions and the main lessons of our whole exercise. We then address the question of their possible implications for development assistance.

I Major Lessons

Three important contributions have resulted from the present endeavour. First, we have proposed a methodology aimed at enabling us to navigate one’s way through the data and information that need to be collected to form a sound institutional diagnosis. Clearly, the approach proposed consists of two major phases: an exploratory phase during which international databases are mobilised and local expert opinions collected, and an in-depth analysis phase intended to probe into strategic sectors or domains where institutional obstacles appear to be especially salient. This meso-level institutional analysis is itself preceded by an economic diagnostic based on the twin concepts of growth engine and structural transformation, the latter being itself rooted in the dual economy framework initially proposed by Arthur Lewis. As we have found, poor countries tend to be characterised by the absence of a clearly identifiable growth engine (Tanzania), an artificial or unsustainable engine (Benin), a misleading engine based on resource rents (Mozambique), or a working engine at risk of misfiring (Bangladesh).

Our second contribution consists of the lessons learned from the institutional analysis: if institutional hurdles and the sectors of the economy where they manifest themselves are not necessarily identical between countries, behind them stand generic issues or problem areas. The identification of these issues provides us with a privileged instrument to penetrate into the institutional territory of a country, more particularly into its parts that matter for long-term development. Our attention has thus been drawn to institutional aspects that have been largely bypassed or generally under-estimated by economists and international experts to this date: the role of initial conditions with special attention to geographic and demographic conditions; the quality of political leadership and state autonomy; state capacity (including property rights, and land rights in particular); the nature of state-business relationships; the functioning of the judiciary; bureaucratic failures with special attention to coordination problems between its various parts; and the role of informal exchanges, including those mobilised for illegal transactions.

Moreover, crucial interlinkages between some of these core problems have been brought to light. For example, state capacity failures are often better understood when they are explicitly related to the political regime. This is typically the case when teacher absenteeism is caused by a political system of clientelism dominated by quasi-feudal local elites. Or, poor coordination between key administrative departments is more likely to be serious and difficult to remedy in contexts where these departments are viewed as political fiefs by the incumbent ministers and their close circles. To take a final example, formal property rights cannot be effectively enforced if the judiciary system is dysfunctional and, in particular, if magistrates and judges are prone to corruption. In the latter instance, indeed, cash notes carry more value than a land title.

As our third contribution, we have proposed the early development experiences of two East Asian tiger countries, South Korea and Taiwan, as useful benchmarks for putting the findings emerging from our four case studies in a broader perspective. The idea was not to use South Korea and Taiwan as models to be replicated but, more realistically, as guides susceptible of attracting our attention to important institutional dimensions of development. The generic issues of the role of initial conditions, the quality of political leadership and state autonomy, and state capacity have thus come out as key entry points into the Taiwanese and South Korean experiences of successful development. Furthermore, the critical importance of state-business relationships, which came out of our four intensive country case studies, has been fully confirmed by our review of South Korea and Taiwan at the time they were more or less at the same development level as these four countries.

There are several important and more specific lessons to draw from the comparison of the development experiences of our 4+2 countries.

Countries may have their own path to development. The fact that countries as apparently similar as Taiwan and South Korea, two showcases of developmental states, exhibited some significant differences in their mode of structural transformation, further reinforces this point. In particular, while South Korea’s growth has largely rested on the dynamism of large firms, small and medium-scale enterprises have played a significant role in the economic development of Taiwan. On the other hand, if the comparison is made between these two countries and Western Europe, we can confirm the idea of Gerschenkron (Reference Gerschenkron1962), according to which the role of the state increases with the distance between the take-off time of a country and the take-off time of pioneer countries. Relatedly, authoritarian regimes have been, and continue to be, a frequent occurrence in presently developing countries, whether in Africa, Asia, or Latin America. It is therefore not surprising that we have found them repeatedly in our small sample of countries (Kérékou and Talon in Benin, Nyerere and Magufuli in Tanzania, the Frelimo in Mozambique, Ershad, Ziaur Rahman, and the two begums in Bangladesh).

In most cases, however, strong leaders have not been to the task, essentially because they failed to see the critical role of private incentives in development and, in some cases, to check corruption in the ruling clique. Those criticisms cannot be levelled against South Korea and Taiwan. In spite of their having many trappings of a socialist state, including economic planning and the centralisation of politics in the form of a single party rule, their political leaders understood that sustained growth cannot be achieved if sufficient space is not created for market forces to exert their competitive pressures. On the contrary, they have exhibited great skills in mobilising private initiative and devising performance-based contracts in their relations to the private sector. It is probably not coincidental, therefore, that East Asian past leaders have inspired the present regimes of Presidents Kagame in Rwanda and Talon in Benin, although in the latter instance the profile of the autocrat somewhat differs from those of his counterparts in Taiwan and South Korea at the time of independence. The short-lived regime of President Magufuli in Tanzania comes as another possible illustration of the authoritarian way to development.

The quality of political leadership is clearly a major factor of successful long-term development. A developing country needs effective leaders able to conduct its destiny with the help of a team of technocrats in charge of setting economic priorities. This implies the existence of an autonomous state immune to pressures from private lobbies and to large-scale corruption, as well as its backing by an efficient bureaucracy. Moreover, it is essential that such leaders prevail over the whole critical period during which the basis of the economy’s structural transformation is laid, which implies that the ‘good’ leader remains a sufficiently long time in power or is followed by successors endowed with similar qualities. This condition has been fulfilled in Taiwan and in South Korea, although in the latter case the country had to wait for the demise of the corrupt leader who seized power at the time of independence. In some other instances, the inverse, less favourable scenario unfolded with ‘bad’ leaders succeeding ‘good’ ones. Thus, in Pakistan, the early rule of Ayub Khan held the promise of an effective state-directed development, yet was succeeded by regimes which, starting with Bhutto and continuing with Zia ul-Haq and Pervez Musharaff, reformed the country’s institutions in a way that deeply changed the rules of the political game. In short, they destroyed the autonomy of the state and the independence of the bureaucracy by encouraging the politicisation of the administration and the subordination of civil servants to elected politicians. Less calamitous is the experience of Indonesia where President Suharto, a strongman who abolished politics and co-opted the army, gave a powerful and consistent impulse to the modernisation of the country during the long period of his rule (no less than three decades). It is on the basis of clear priorities decided at the top that business cronies were allowed to carry out the national plan. By contrast, President Joko Widodo (popularly known as Jokowi), who did not hail from the army or the country’s elites, ended up frustrating the many hopes that his election raised among ordinary citizens. Mostly adept at entrenching his power through vast co-option of potential rivals, he let his cronies set the priorities. Damage has been more limited than in Pakistan because the engine of sustainable growth had been set into motion before his advent to power.

As we also learn from the East Asian experience, including post-1978 mainland China, an effective state-directed development is grounded not only in a successful industrial policy but also in sizeable investments intended for rapidly expanding the country’s infrastructure and its human capital. In doing so, major attention needs to be paid to quality factors. This is especially evident in the case of education where performance must be measured not so much by admission or graduation rates as by learning outcomes.

II Implications for Development Assistance

Our review of the four countries selected for intensive study has revealed that lack of insulation of the state from private business interests, faulty economics, and poor state capacity have largely contributed to disappointing development performances. Foreign assistance can help enhance the quality of the bureaucracy and to better manage the economy, but only in so far as bureaucratic effectiveness and quality of economic management do not depend on political factors. In reality, however, many institutional failures are traceable to the way political forces in a country play out.

Because almost all country diagnoses have so far ignored this basic truth, it is no surprise that hopes placed by Western donors in the development prospects of particular countries have been frequently dashed. Upon careful look, disappointment has been caused not so much by the sheer ignorance of politics as by the influence of rather naïve views underestimating the complexity of political economy mechanisms. This has been recently testified by the disastrous experiences of Western aid in countries such as Haiti, Somalia, Libya, Iraq, and Afghanistan. Moreover, because political patterns are influenced by the historical path of a country, its geopolitical situation, and particular aspects of its social fabric and culture (think of the role of clans and other traditional entities, of immigrant entrepreneurs, religious movements, regional and ethnic divisions, and so forth), any political economy analysis needs to be country specific. This does not mean that there are no common features between certain countries, but they should not be assumed a priori on the basis of superficial examination, and specific features should be investigated.

A first lesson is therefore that donors need to have a good knowledge of the countries it deals with, particularly with respect to the dimensions that our study has shed light on. Examples abound where interventions were made without sufficient information about potential backlash effects, without proper cognisance of impinging informal institutions (e.g., land tenure rules, contract enforcement mechanisms, exchange and credit relations) and their resilience, or without due awareness of the complex interests at stake. Second, since politics strongly affects development results, a donor cannot ignore the role of governance, and this implies that aid allocation decisions must necessarily be made on the basis of a needs-governance trade-off. In addition, the donor must decide whether the aid awarded is going to be effectively monitored with inevitable sanctions when fraud is detected. We have seen that in some cases, donors have shown a lot of laxity in the presence of a series of well-published scandals, presumably because stopping aid also involves costs for the donor country (due to fixed costs of establishing the aid link, to the harm caused to geopolitical interests, or to the ensuing losses for its own business firms). Also, their readiness to use ad hoc measures to bypass the implications of governance indicators is to be questioned.

As shown in Bourguignon and Platteau (Reference Bourguignon and Platteau2021a, Reference Bourguignon and Platteau2021b), the advantage of using donor monitoring is that it increases the chance of poor and badly governed countries to be eligible for aid, provided, however, that the initial gap between their governance quality and the quality observed in other poor but better-governed countries is not too large. Budget aid, grounded in the idea of partnership between donor and recipient, is evidently not recommended when conditionality needs to be imposed by the former on the latter. When a donor deals with poor and badly governed countries whose populations cry for external assistance, project aid, and to a lesser extent programme aid, seem advisable. The reason is that such aid modalities are more easily monitored, in part because they may be implemented by NGOs or private companies contracted by the donors. An extreme case is the mode of intervention favoured by China when financing and building infrastructural facilities in Africa. Since monitoring intensity may be tailored to the specific governance situation of each recipient country, recourse to project aid need not be systematic: appropriate for so-called fragile countries, it may be largely dispensed with in poor and better-governed countries.

Devarajan and Khemani (Reference Devarajan and Khemani2016: 21) make the point that, when government failures are the result of politics, priority ought to be given to aid modalities for which citizens’ ability to hold the state to account is maximised. In short, aid should primarily aim at promoting citizen engagement and transparency (2016: 21). In the same line, Wade (Reference Wade1985: 488) stresses the need to strengthen the rules of public disclosure or accountability to curtail corruption. This prescription raises the moot question as to whether aid should aim to improve governance directly or, rather, to lay out the infrastructural basis of long-term growth, promote better education, health and technical training, and reduce poverty. We submit that the latter option should be followed but strict conditions and rigorous monitoring should be attached to the disbursement of aid so as to make it as effective as possible: in this way, aid-related governance rather than general governance can be improved. Moreover, whenever people’s participation and self-assertion are an important prerequisite of success, as in many rural development programmes, the donors must encourage the underlying processes. In many cases, this may require that on-the-ground interventions are entrusted to non-government organisations, or similar types of associations, both local and international.

We thus believe that general governance in recipient countries is not susceptible to being exogenously modified, whether by demanding public transparency or otherwise. Placing great hope in the ability of foreign actors to initiate changes in governance is bound to lead to disillusionment, and it may even happen that such attempts will spark nasty backlash effects. As already suggested, the more realistic solution consists of applying aid allocation rules based on comparative country performances weighed down by considerations of intensity of needs and coupled with a serious monitoring of the aid use. On the other hand, as the Taiwanese historical experience testifies, improvements in state accountability and the promotion of a civil society evolve gradually and are the outcome of development rather than its prerequisite. Furthermore, if economic liberalisation was largely undertaken under foreign pressures, political liberalisation was to a great extent an endogenous process, in which even the authorities eventually participated. The problem is of course more complicated when political leadership is lacking, say because the clique in power is corrupt or the ruler has unrealistic views regarding the country’s growth potential. It is in this sort of especially hard circumstances, however, that effectiveness in aid use should arguably be a major preoccupation of the donor community.

There is a snag, however. Endogenous or co-evolving progress toward political liberalisation is more difficult today than it was at the time Taiwan and South Korea developed. The reason is that globalisation and the accompanying exposure of elites and middle-class people to modes of living and values in developed countries, including democratic aspirations, cause their preferences to change more quickly than ever before. As a result, restrictions to freedoms have become less acceptable to people, the educated classes in particular, thereby requiring a greater measure of people’s indoctrination, internal repression and isolation from international influences. The examples of Russia and China come to mind here. This reality thus points to a tragic dilemma for which no satisfactory resolution seems to exist. And ominous consequences for the international order also follow.