Based on field research conducted in Burundi, I explore the implementation gap between good governance policies and informal practices by street-level public employees (tax collectors, local accountants). These practices include bribery issues or cronyism among relatives and community members. Through corruption analysis, my work shows it is possible to better understand the relationship between state and society, state functioning, and the importance of social norms in public employees’ behaviour. Moreover, such analysis on corruption may question the role of development aid, in particular the support provided to good governance and decentralisation programmes.

Governing party’s role

Corruption in Burundi is systemic. Corrupt practices persist mainly due to insufficient monitoring, as controllers (the persons tasked with challenging corruption at the national and local levels) also reproduce the same dynamic of corruption under scrutiny. It would be dangerous, for reasons I explain below, for controllers to sanction influential local leaders. It is much more convenient for them to set up informal arrangements, which in turn generate more corruption.

The current leading political party in Burundi is the National Council for the Defense of Democracy–Forces for the Defense of Democracy (CNDD-FDD). Its grip on power, and the great risk for whistleblowers posed by their peers and security services, together generate a kind of immunity for public agents and local leaders linked to the party. In particular, the CNDD-FDD’s omnipresence in public servants’ decision-making inhibits their willingness to oppose corrupt practices. Public agents consider loyalty to social rules and the party to be more important (or safer) than respecting formal rules.

For instance, when an accountant at the commune level (Burundi’s second largest administrative division) was asked why he covered up the mayor’s malpractices, he answered, ‘It was too risky to oppose the administrator, whose influence on and support by the party would harm anyone who tried to denounce this situation’.

Social networks versus formal rules

Burundi citizens have little confidence in the state and this encourages them to accept bribery in the face of authority. As the government has been and still is discredited in the eyes of the public, loyalty leans more towards social networks than public institutions.

For example, a case of corruption was reported concerning a local market, Kiziguro, with a goat-keeper bribing a tax collector instead of paying market fees for the sale of a goat. A large section of the crowd supported the alleged offender when he said publicly, ‘Theft in this country is not new, and it is not in Kiziguro that we should start penalising corrupt people’.

This illustrates the extent to which accessing public services and interacting with public employees is not necessarily driven by notions of formal rights and entitlements. On the contrary, network building with public employees helps to mitigate uncertainty and gain access to public services. Accordingly, formal rules are socially embedded. This means that public employees use their social inclusion (their identity or status) and their formal authority to perform their duties and make money from their prerogatives.

Of course, petty corruption is very much determined by socio-economic conditions. For example, tax collectors’ salaries are so low that they argue they have no other choice but to disregard rules in order to survive. Besides this economic rationale, some informal practices are supported by power games. For example, an inspector at the province level reported embezzlement by a local administrator. Yet, the local administrator was then appointed as a new governor and head of provincial inspection; he quickly quelled any challenges to his integrity by covering up the report and threatening the inspector’s family.

Solidarity ties among public agents are another key enabling factor for corrupt practices. Tax collectors sharing informal earnings among team members is a case example. Other such factors are related to social ties between public agents and their community. These highlight the different kinds of legitimacy behind corrupt practices.

Tax enforcement investigation

An interesting task is then trying to understand how this search for private gain is framed at the local level by social norms. The social structure offers room for manoeuvre, but also constraints — whereby public employees must follow practices and solidarity ties related to their socio-professional environment. An investigation into tax enforcement provides a basis to measure the discrepancy between formal good governance norms and informal strategies developed by local civil servants and officials.

Within communities supported by a good governance programme in Cibitoke province, reforms attempted to improve the tax collection process. The aim was to increase the municipal budget and to improve relations between the state and citizens.

One of the approaches was to replace ‘hills’ representatives’, who were considered to be corrupt, by professional tax collectors. Hills’ representatives represent urban or rural neighbourhoods, the smallest sub-division in Burundi. This makes them the elected leaders nearest to the population. Despite amendments to the existing legal and administrative frameworks, petty corruption continued. What is more, these reforms simultaneously reduced the willingness of hills’ representatives to facilitate tax collection. They had less interest now to support the tax system without the additional gains they were making from bribes. Within a short time, tax revenue diminished as tax avoidance increased.

This tax avoidance was directly related to taxpayers’ mistrust of local authorities, mainly because the hills’ representatives were no longer involved in collecting taxes. It was clear that they had been the ones providing legitimacy to tax collection, embedding the process to social ties. Moreover, power games emerged between unhappy, newly disenfranchised hills’ representatives and the professional tax collectors: for instance, when the hills’ representatives did not facilitate tax processes or no longer ensured the safety of tax collectors (a job that can be dangerous in Burundi).

Ultimately, the good governance programme supporting tax reforms had altered a delicate equilibrium on tax consent, based on the hill representatives’ legitimacy and the informal economy — itself propped up by petty corruption.

A way forward

Corrupt behaviour is a social phenomenon — far from its formal definition, which generally refers only to private gains and not social conditioning. Yet recognising this could lead to what Jackson and Köbis term ‘entry points for intervention’, in which more relevant and localised anti-corruption reforms are envisaged, ones in which social and political pressures are addressed. For example, Bicchieri’s methodology is helpful to identify and assess the rationale behind behaviours, by looking at people’s empirical and normative expectations within a social group. This in turn may be useful to measure and to modify the influence of social norms and their support for corrupt practices.

This investigation underlines the complexity of transferring good governance norms in an environment where public employees and citizens rely on an informal patron–client relationship to facilitate access to public services. Programmes that aim to transfer good governance norms must consider how formal rules are woven into informal practices.