ECONOMISTS know about what they call ‘market failure’ as well as ‘government failure’ in a situation where neither institution as a whole, for a number of well-recognised reasons, has been able to deliver development.

Issues of coordination, asymmetric information and inequality are said to affect both avenues of development. Such theoretical understanding and practical experience gave rise to the need for ‘civil society’ to emerge and resolve such failures and offer itself as a developmental alternative, and a much better one than either.

Assumptions have been made that groups of people, ranging from village communities, urban neighbourhood councils, school associations, and water user groups, work for the common interest of their collective members, benefiting all.

A sense of egalitarianism has seeped into the general framing of civil society’s efforts to provide development, and this has given rise to a vast global movement of self-help, or NGO-led development models, where NGOs and community-based organisations (CBOs) have replaced governments or become contractors for them.

The surprising thing is that there has been very little empirical evidence, until now, that is, which examines the contributions of civil society to development. Conventional wisdom endorses the broad, feel-good sentiment, that civil society does good, a great deal of good, for communities, and that it is a far better alternative to either the government or the market.

In a recent book produced by two economists at the World Bank, the model of civil society development has been empirically challenged. Even though the World Bank has allocated $85bn to local participatory development over the last decade, there has been little detailed investigation as to whether this participatory model of development works well.

Ghazala Mansuri, a former graduate of Karachi University and her colleague Vijayendra Rao in Washington have looked at over 500 empirical studies of participatory development interventions, and their findings are a loud wake-up call for all civil society developmental enthusiasts.

The wide-ranging notion of civil society needs to be bifurcated into at least two broad categories: advocacy-related, organic, participatory movements (such as those which overthrew military dictatorships or the Soviet-backed regimes of the 1980s leading to widespread political and social change); and what the authors call ‘induced participatory interventions’ (largely working for developmentalist outcomes and goals). It is this latter, the so-called NGO or community-based world, which comes under careful scrutiny by the two economists, where ‘fads, rather than analysis, tend to drive policy decisions on participatory development’.

They find that many of the problems which afflict government and market also affect civil society initiatives, and that inducing civic empowerment may not be a better choice than a market-based strategy or even one which strengthens the role of centralising bureaucrats. Groups do not always work collectively, as one has assumed; there is a great deal of differentiation within such so-called communities, where differences of class, gender and status affect outcomes and lead to elite capture.

Their research shows that participants in civic activities “tend to be wealthier, more educated, of higher social status, male, and more politically connected than non-participants”. Given such existing divisions in all communities, they feel that there is little “evidence that induced participation builds long-lasting cohesion, even at the community level”, and many existing social divisions are actually reinforced. The poor benefit less from participatory processes than do the better off and, despite their focus on poverty alleviation, “community-based development efforts have had a limited impact on income poverty”.

One of the assumptions of this participatory development, whether of the NGO kind or of decentralisation, is that it improves development outcomes and that there is likely to be equitable development. Looking at a wide spectrum of initiatives, the authors find again, that decentralised projects lead to local capture by the elite and the better-off, and often projects benefit the better off rather than the poorer communities for whom they were initially intended. Especially women and minorities, who are excluded anyway, continue to be discriminated against in such so-called participatory, localised, decentralised, developmental initiatives.

What is not surprising is that civil society has failures embedded in its developmental model just as much as government and markets do. What is surprising is that there has been little careful evaluation of civil society efforts and that they have been promoted by donors, aid-giving foundations, communities, even governments.

One of the key findings of this research is not surprising: that repairing civic failures requires reducing social inequalities.

The state, rather than the market or civil society, is perhaps best equipped to address such inequalities and, importantly, as we have seen around the globe, the state does not need to be democratic to do so. The evidence provided by these authors “overwhelmingly suggests that effective community-based interventions have to be implemented in conjunction with a responsive state”.

The writer is a political economist.