2. On institutional copying

New institutional economics regards institutions as durable systems of humanly devised constraints or social rules (Hodgson, Reference Hodgson2006: 13; North, Reference North1990: 3). There is a consensus today that institutions are fundamental determinants of economic growth (Acemoglu et al., Reference Acemoglu, Johnson and Robinson2001; Rodrik et al., Reference Rodrik, Subramanian and Trebbi2004; critically, Chang, Reference Chang2011). Reforms in developing countries often involve an element of emulating institutions from developed countries. However, such institutional copying often fails (Grindle, Reference Grindle2004; Rodrik, Reference Rodrik2008). In this section I review what we know about institutional transfers with the aim of extending our understanding of how informal norms are transferred.

Many institutional economists would agree that institutional transfers fail when transferred foreign formal institutions conflict with local informal norms (Boettke et al., Reference Boettke, Coyne and Leeson2008).Footnote ¹⁰ A growing field of legal transplants or (more generally) institutional transplantsFootnote ¹¹ has emerged around this argument. Couyoumdjian (Reference Couyoumdjian2012) gives an overview on this literature. In an empirical study on transplanted law Berkowitz et al. (Reference Berkowitz, Pistor and Richard2003) argue that imported foreign law was largely ineffective, unless it was intentionally adapted to local law. Outside the rare cases in which countries share a close common and legal history (e.g. Scandinavian countries), the lack of adaptation renders the legal transfer as dead law – an outcome that authors call the ‘transplant effect’. The work of Berkowitz and colleagues is an important contribution to the field, even though their empirical model has been criticized for being based on a simple categorization (Daniels et al., Reference Daniels, Trebilcock and Carson2011). Furthermore, the authors offer little indication of why some countries were better in adapting than others or of who was in charge of institutional adaptation (Seidler, Reference Seidler2014). More recently, development economists have argued that the transplant effect may well be an accepted outcome in the developing world, when governments are content with mimicking the outside appearance of institutional reforms in exchange for continued financial aid (Krause, Reference Krause2013; Pritchett et al., Reference Pritchett, Woolcock and Andrews2013).Footnote ¹² Can the problem of conflicting local institutions be overcome? Institutional literature offers two arguments. This paper will present a third alternative in this section.

First, the deliberate and top-down change of existing informal institutions is considered difficult but possible. The assumption is that newly introduced formal institutions can inculcate values into people living under them. When an imported institution creates regularities of behaviour among individuals, they begin to internalize the values embodied in the new regime (Chang and Evans, Reference Chang, Evans, De Paula and Dymski2005: 103; Hodgson, Reference Hodgson2006: 7). Ex-post enforcement is a central factor. For example, curbing Jim Crow practices in the USA in the 20^th century needed strong federal enforcement by the new civil rights legislation (Helmke and Levitsky, Reference Helmke and Levitsky2004: 732). Post-reform enforcement costs are substantial unless marginalized groups exist who benefit from the new regime (Aldashev et al., Reference Aldashev, Chaara, Platteau and Wahhaj2012) or there is a change in the underlying conditions under which the informal institutions emerged. For example, the purchases of army commissions ended in Britain when new technology improved the monitoring of officers in battles (Allen, Reference Allen2012). In a similar manner, the end of Chinese foot binding took place when European influence had created new education and employment opportunities for women in the early 20^th century, which in turn shifted the marriage market in China (Mackie, Reference Mackie1996). ‘Sharp breaks from established procedures’ such as crises or other critical junctures can be manifestations of a rapidly changing status quo facilitating the deliberate change of informal institutions (Williamson, Reference Williamson2000: 598).

A second option is to adapt the imported formal institution skilfully to local informal institutions in the receiving country. Rodrik (Reference Rodrik2008) and Grindle (Reference Grindle2004) favour such tailored ‘second-best’ institutions over what Evans (Reference Evans2004) calls ‘institutional monocropping’, i.e. the imposition of imported institutional blue-prints. The authors admit that adapting is not widely practised by international development agencies today. In theory, the resulting hybrid institutions should lower the enforcement costs of new regimes as local actors recognize the familiar elements therein and the new elements are not found too contradicting with how things have been done before. Unlike option one (a top-down change of informal institutions), most institutional adaptation takes place before the new institutional regime enters into force. Kurkchiyan (Reference Kurkchiyan2009) has investigated how a media self-regulatory body (an institutional copy of the BBC) was established in two different and geographically distant regions in Russia in 2002. While British media experts provided information on the model type, Russian experts were in the driving seats of establishing media self-regulatory agencies. Because of a distinct legal culture in Russia the resulting agencies had little resemblance with the British original but fulfilled a similar function. Japan is the more classical example of successful nationwide institutional adaptation. During the Meiji period, skilled Japanese bureaucrats imported European institutions and created well adapted copies (Westney, Reference Westney1987).

Is adaptation then the dominant strategy? Mukand and Rodrik (Reference Mukand and Rodrik2005) present a formal model in which states face a trade-off in the costs and benefits between adaptation and blueprint copying. The former strategy risks corruption when political interests take over, while pure imitation may result in a poor fit and end up as dead law. Grajzl and Dimitrova-Grajzl (Reference Grajzl and Dimitrova-Grajzl2009) describe a similar model in which developing countries make a choice between copying from abroad or developing indigenous institutions. In both formal models the human capital needed to adapt foreign institutions is available at no cost. However, in developing countries, skilled bureaucrats, who need to be knowledgeable in both the complex formal rules of foreign states and the prevailing informal norms in the receiving country, are often extremely scarce. I argue that blueprint imitation takes place in developing countries, precisely because there are so few well-trained intermediaries required for the adaptation process. From this perspective, developing countries have no choice and some degree of blueprint copying will be the default strategy. With little means of enforcing the new regimes, reforms in developing countries are likely to create dead law and corruption.

A look at the related and rich literature on diffusion supports this argument. Diffusion is understood as the transfer of innovations among organizations and corporations.Footnote ¹³ Here, the lack of fit owing to dissimilar organizational cultures poses an analogous problem. In international business however, efficiently set wages attract suitably trained workers. Hence, the adaptation or ‘re-invention’ of transferred innovations has become the standard response in this literature (Rogers, Reference Rogers2003: 17; Stone, Reference Stone2012: 483) and studies have focused on the speed or the extent of adaptation (Ansari et al., Reference Ansari, Fiss and Zajac2010). More recently, the diffusion literature has turned its attention to the transfer of policies into developing countries. Questions here are again very similar to those debated in the institutional copying literature. Among others, agency and structure (Marsh and Sharman, Reference Marsh and Sharman2009: 274–5) or the transaction costs of institutional adjustment of foreign policies (Benson and Jordan, Reference Benson and Jordan2011: 372–3) have become a major focus. Consequently, the success of policy transfer is to some degree seen as a function of available human capital and unlike in the institutional literature, there is a rich debate on actors involved in policy transfer (Benson and Jordan, Reference Benson and Jordan2011: 368–9). In China or inside the EU, where skilled bureaucrats are available, the adaptation practices (‘translation’ or ‘bricolage’) have taken a further step away from the original policy when scraps of imported policies are pieced together to create a new hybrid policy (de Jong, Reference De Jong2013; Stone, Reference Stone2017: 60; Thelen, Reference Thelen, Mahoney and Rueschemeyer2002: 227; Zweynert, Reference Zweynert2009: 353).

How much institutional bricolage took place in decolonizing Africa? The British had no central policy of deliberate adaptation of imported institutions to local conditions. Instead there was a general interest in maintaining high, i.e. British, standards in practically everything from localization to housing construction (Burr, Reference Burr1985:130; Tanganyika, Reference Tanganyika1954:17). Institutional adaptation took place infrequently at best.Footnote ¹⁴ Post independence, enforcement was low across Africa.

Under these conditions a third argument may explain variations in the success of institutional transfer. The conflict of imported formal institutions with local informal norms could have been mediated by the concurrent transfer of those informal institutions on which the foreign formal regime rested in its country of origin. In a similar manner, the diffusion literature has very recently started to distinguish between hard and soft transfer (Stone, Reference Stone2017: 8). Institutional theory informs us that formal and informal institutions rarely function alone, but complement each other to form the institutional framework of a society (North, Reference North1994: 366), and that piecemeal copying of selected institutions can be expected to fail (Roland, Reference Roland2004: 113). Also, the congruence of informal and formal institutions matters to economic performance, while formal institutions alone are less likely to promote economic growth (Williamson, Reference Williamson2009). To my knowledge searching for hard and soft transfers in the colonial past has not been done before. Berger (Reference Berger2009) shows in the example of colonial Nigeria how tax collection increased bureaucratic capacity with long-lasting effects on local governance. Here, the mechanism also runs through the human capital in local colonial governments and the author rightly argues that their effectiveness relied more on the competence of bureaucrats than on the law in place. Very unlike the more influential literature on long-term effects of colonization (for example Acemoglu et al., Reference Acemoglu, Johnson and Robinson2001; Dell, Reference Dell2010), this paper's focus on the administrative capacity at the end of colonial rule offers more proximate causal relations.

3. Institutional copying on the example of the Public Service Commission

This paper's analysis follows the third argument, which proposes that importing foreign institutions falls into two processes: copying the formal law and transferring its informal foundations with it. The two processes have different needs in two categories: actors and time. The introduction of the Public Service Commission (or simply ‘the Commission’) into British colonies serves as the object of investigation. Commissions had the task to ensure the political neutrality of civil servants by preventing nepotism and cronyism in particular in the areas of appointments, transfers and promotions. All African colonies created an initial version of the Commission in the years after 1945. The first Commissions merely advised the governor on questions of appointments and disciplinary measures. In negotiations on self-government the governors ceded powers and the British made it their policy to raise the status of the Commissions to independent bodies with executive authority in all matters related with appointments, transfers and discipline of civil servants. In African colonies, this took place about five years before independence was granted (Adu, Reference Adu1965: 129–45; Hatchard et al., Reference Hatchard, Ndulo and Slinn2004: 72; Lee, Reference Lee1967: 205).

It is not clear whether an equal understanding of the notion of political neutrality was shared by all individuals working inside the colonial administrations. Even within Colonial Office itself, a certain degree of patronage had been practised until the 1930s (Kirk-Greene, Reference Kirk-Greene2000: 129, 216–7). One can assume that the formal rules safeguarding the political neutrality of civil servants were foreign to colonial Africa and in many cases incompatible with local institutions shaping executive authority (Englebert, Reference Englebert2000). In consequence, the intention behind the processes and procedures related with the Commissions may not have been recognizable and may have been considered overly burdensome to British and local staff. While British colonial officers grudgingly accepted the burden of new, complex bureaucratic procedures and the paperwork that came with them (Lovering, Reference Lovering2010: 6), local officers from the colonies regarded them as ‘straight importations of structures which had proved workable or effective in the United Kingdom and which could work in Africa so long as there were British officers to man them’ (Adu, Reference Adu1965: 116).

The Colonial Office in London recognized that in the lack of ‘a body of convention and tradition to supplement its work’ the Commissions needed wider functions than their counterpart in the UK, the Civil Service Commission (Colonial Office, 1962: 2). London was also aware that the Commission was a foreign transplant and likely to be opposed by local staff and local politicians. A handbook advised overseas officers on how Commissions were best established in the colonies and what problems were to be expected.Footnote ¹⁵ It also contained specimen regulations; i.e. legal text to be used as a blueprint for equivalent colonial provisions (Colonial Office, 1962).

The first step of the analysis assesses variations in the copying of the formal institution. Before the British granted independence, the Commissions were established as fully independent bodies inside the public services and their powers were preserved in the constitutions of the newly independent states. Column 1 in Table 1 compares the legal provisions in these nine constitutions with the colonial blueprint along the following three legal requirements laid out in the ‘Specimen Public Service Commission Regulations’ (Colonial Office, 1962: Appendix A).Footnote ¹⁶

1. 1. The Commission needed to be established as independent body apart from other bodies of the civil services.

2. 2. The Commission had a minimum of three full-time members to be appointed by the successor of the British governor after independence – i.e. the parliament, the prime minister or the president.

3. 3. Third, the Commission's powers included all appointments, promotions (transfers) and disciplinary measures of all civil servants below ministerial ranks. Judicial service officers, who had their own Service Commission, were excluded.

Table 1:

Formal law and legal practice today related with the political neutrality of civil servants

Sources: Column 1: Constitutions of African countries (see Appendix 2a); columns 2–5: for African Integrity Indicators (2013–17) see Aii.globalintegrity.org (2017).

Column 1 shows that at independence, only the constitutions of Sudan and Swaziland deviated from the standard text. The changes were intentional. They constituted a degree of institutional adaptation or bricolage that reflected the political and legal institutions in place at the time of the copying. In Swaziland, the Public Service Commission had already been established in 1963 with advisory powers only. On the request of the Swazi king this was simply enshrined in the 1968 constitution (Proctor, Reference Proctor1973). In Sudan, the geopolitical background tensions behind Sudan's independence weakened the British position as the country prepared for independence (Sconyers, Reference Sconyers1988). The constitutional provisions of the other colonies in the sample were almost verbatim copies of the corresponding paragraphs in the Colonial Office Handbook (1962: Appendix A).

Columns 2 to 5 in Table 1 fast-forward to the 21^st century. They use the Africa integrity indicators (Aii), a set of indicators comparing formal legal text with the extent to which the law is effectively practised in a country. Since independence, legal reforms have changed and in many cases weakened the Commissions' powers across former African colonies. Column 2 tests whether there is formal legislation in the spirit of the British Public Service Commission today. A ‘Yes’ is assigned by the Aii when at least two of the following three conditions are met:

1. 1. the law prohibits conflicts of interest, nepotism, cronyism, and patronage (at least two of these offences must be prohibited),

2. 2. the law applies to all branches of government, including the civil service,

3. 3. the law mandates mechanisms such as competitive recruitment and promotion procedures, safeguards against arbitrary disciplinary actions and dismissal, and recusal procedures.

Two out of nine former colonies do not pass the test. Ghana amended its constitution two years after independence introducing ‘special cases’ of appointments outside the Commission's strict mandate (Harvey, Reference Harvey1966: 57). Ghana's latest constitution (of 1992) and the Civil Service Act (of 1993) both address conflicts of interest but issues of cronyism, nepotism and patronage have been dropped. Swaziland today has no corresponding formal or constitutional law at all. At the other extreme, Gambia and Sierra Leone have kept almost verbatim copies of the British blueprint in their constitutions.

Finally, columns 3 and 4 assess the legal practice of the formal law in place. The figures are averages of the African integrity indicators (Aii) between 2013 and 2017.Footnote ¹⁷ This is in fact a measure of the transplant effect – the degree to which an imported foreign law has become dead law decades later (Berkowitz et al., Reference Berkowitz, Pistor and Richard2003). Whereas columns 1 and 2 report on formal legal evidence, columns 3 and 4 give perception-based indicators. Here, the Aii assign the maximum score of 100 points when the legal provisions laid out in column 2 are effectively practised. Two separate indicators assess this. In column 3 a score of 100 is earned when both of the following conditions are met:

1. 1. Civil servants are disciplined/removed/transferred only through due process by a peer panel or oversight body.

2. 2. Civil servants are not removed when a new administration takes power.

In column 4 a score of 100 is assigned when all the following conditions are met:

1. 1. Appointments to the civil service are made on a merit-based system.

2. 2. Individuals appointed are free of conflicts of interest due to personal loyalties, family connections, political party affiliations or other biases.

3. 3. Performance evaluations are based on standard benchmarks.

Because columns 3 and 4 both assess the legal practice of the same body of formal law, it makes sense to build an aggregate value of the legal practice in place per country. Column 5 shows that there is considerable variation in this measure. The formal legal provisions do not appear to be an influential determinant. Ghana ranks second in terms of legal practice but has an insufficient formal law in place. Gambia has preserved the colonial law in its constitution but comes out weakest in terms of legal practice.

How much of the variation in the legal practice can be explained by the presence of British officers? Figure 3 reports on the connection between the legal practice of preventing political influence in the civil service with the paper's measure for the presence British civil servants from the introduction. The vertical axis shows the average Aii scores of the legal practice for each country (i.e. column 5 in Table 1). The horizontal axis displays the number of years in which the proportion of British officers in the civil service remained between 25 and 75% in the civil service.

Figure 3.

Effectively practised prevention of political influence in civil services and the presence of British civil servants in former colonies

Sources: Territorial staff lists and secondary sources (see Appendix 2b and 2c); for African Integrity Indicators (2013–17) see Aii.globalintegrity.org (2017).

The scatterplot indicates a positive relationship between the share of British officers in the civil services after independence and the extent to which formal provisions protecting civil servants from political interference are practised effectively today. Again, the association holds as long as the share of the British officers remains above 25%. Very similar patterns emerge regardless of whether British officers were in the majority (between 50% and 75%) of the civil service or in the minority (between 25% and 50%).

4. The mechanism and identification

A series of telephone interviews with members of the Overseas Service Pensioners’ Association in the UK lends supports to the argument around the transfer of conventions into African civil services. The interviews followed a mixed methods questionnaire combining biographical questions and quantifiable ones. Question 53 from the interviews comes closest to asking about the mechanism behind the effect of their presence.Footnote ¹⁸ The results presented here are based on the responses of 45 officers who remained in service after independence. They were asked two questions:

1. 1. ‘In your opinion, how relevant was the presence of British staff after independence for the local staff when they worked with imported British structures?’

2. 2. ‘What were your own experiences?’

While 84% considered the presence of British staff relevant, the officers’ experiences are clustered around themes. The majority (25 out of 45 officers) thought that giving advice and being around mattered most. Personal accounts differ, but in particular the first years after independence were perceived as a continuation of colonial work routines by these officers. This is plausible. After independence, former British officers were no longer in control, but many enjoyed access to the ministers, politicians and heads of departments who had been working below them (Selassie, Reference Selassie1974: 193). Nine officers (20%) thought that their presence helped because it bought time needed for recruitment and training of local successors. Five officers saw themselves as actively maintaining British standards after independence. An education officer in Ghana for example felt that in departments where British officers were already in lower numbers one had to conform, and corrupt practices were more likely to occur.

Alternative explanations

Explaining the post-independence governance of former colonies by the extended presence of expatriate civil servants after independence is a new line of argument. An influential part of the existing literature stresses a more traditional understanding of local human capital. Here, European education in the colonies enabled economic growth and with it better institutions (Bolt and Bezemer, Reference Bolt and Bezemer2009; Glaeser et al., Reference Glaeser, La Porta, Lopez-de-Silanes and Shleifer2004; critically Acemoglu et al., Reference Acemoglu, Gallego and Robinson2014). In the context of African colonies, it seems plausible that a large pool of the European-educated local population increased the likelihood of Africans being employed in the higher civil service at an early stage before independence.

Table 2 compares the Aii-based measure for the legal protection of civil service neutrality (column 1) with the figures linked to human capital. Column 2 lists enrolment in European schools in the late 1940s.Footnote ²⁰ Column 3 displays the share of Africans in the colonial services five years ahead of independence.Footnote ²¹ Neither school enrolment nor localization rates five years before independence seem related with the legal practice in 2017. Ghana lives up to its reputation of being a colony with an early investment into education, which is reflected in a large share of local staff in senior positions five years before independence. However Nigeria, where colonial education was a lower priority, was similarly fast in localizing its senior officers.

Table 2:

Government effectiveness 1996, colonial education and the share of local colonial officers 5 years before independence.

Sources: For African Integrity Indicators (2013–17) see Aii.globalintegrity.org (2017); Bolt and Bezemer (Reference Bolt and Bezemer2009), territorial staff lists and secondary sources (see Appendix 2b and 2c).

The share of British staff may simply reflect the viability of African colonies in the sense that territories that found it harder to gain independence had more need of expatriate civil servants. In such cases, the share of local staff five years before independence should be low in colonies with later independence relative to those colonies with an early independence. However, colonies with later independence (Swaziland, Gambia) have shares similar to those of Sudan (1956) or of Sierra Leone (1961). This suggests that that the variation in the presence of British staff set in after independence, when inducement payments in the overseas compensation scheme took effect.

Finally, some colonies may had been considered more promising than others and British officers would have wished to remain in service there – resulting in the relatively high number of expatriates in successful former colonies. A majority of interviewed officers dismissed this idea. Likewise, experts and officers at that time potentially misjudged future development potential. At independence, no one had any hope for Botswana, while Nigeria was considered to have huge potential (Acemoglu et al., Reference Acemoglu, Johnson and Robinson2002; Stolper, Reference Stolper2003).