7. Use retailer, employer, and public works competition ''for the poor''

When extra income is generated by public works, it is spent; if received by the poor, it is spent mainly on food, especially staples. There has been much controversy about the circumstances under which this is best handled by providing wages in kind, especially food-for-work. This book will not review that controversy, but one thing is clear. Reliance on monopolistic food delivery imposes heavy delays and other transactions costs, and possibly price gouging, upon the poor. In Botswana, drought relief worked moderately well in the 1970s because food was supplied through a variety of retailers, always potentially and often actually competitive. However, the programme was substantially more cost-effective in benefiting poor drought victims in the 1980s after the switch from direct food distribution to labour-based relief, because the whole competitive retail system (not just food suppliers) responded to the extra wage demand, with no attempt to "force" food [Dr and Sen 1989: 156].

Even more important than retailer competition for food is employer competition for labour: a market relationship, understood in advance by the programme authorities, between them and private employers. Indeed, while public works employment always implies public (or community or NGO) financial provision, the production, for example of roads or irrigation maintenance, can sometimes be carried out by private employers, selected after competitive tenders to the public sector. Public works schemes normally include paths towards efficiency (e.g. stipulated tasks, rates, and supervision systems) and towards effective poverty reduction (e.g. rules on labour intensity, wage share, etc.). However, given public provision of a set amount of resources for public works, there is no iron law that public production and management will meet the efficiency conditions or the anti-poverty conditions better than private employers facing them as (subsequently monitored) rules for competitive tenders. Both sets of controls are imperfect (and corruptible); which is better, for which works and in which circumstances, is an empirical question.

Lacking provision for tendering to private constructors or work organizers, public works managers certainly face laborious requirements to orientate activities towards efficiency and the reduction of poverty. This is partly because financial and physical controls inevitably involve conflicts between finance departments and executing (line) departments. In private firms the board of directors, or the individual boss, resolves such problems. In publicly managed works programmes, they are in principle resolved at political level, e.g., for Maharashtra's EGS, the district zilla parishad or the lower-level block development office. Yet the conflict between line departments and the revenue department remains, as do high enforcement costs.

Elaborate arrangements are made to deliver the EGS benefits to the recipients. The revenue department must be prepared to provide work on demand but it is the irrigation, forestry and other departments which draw up advance plans for work in groups of villages and then, on instruction, execute individual projects using the EGS labour. Official instruction, informal guidelines, extensive monitoring, unscheduled field visits, vigilance tours by officials at various levels, and the advisory and supervisory role of non-official statutory committees help in delivering the EGS benefits [Dev 1994: 14].

EGS is a successful scheme, though perhaps decreasingly so. Yet, even there, the above inevitably costly and exposed process of public production and management is not obviously better than competitive tendering for private execution (also a risky process). The options should be fully and publicly reviewed, when the scheme is planned and at intervals thereafter, separately for various types of works and local circumstances. It is easier to keep this review process, and public works as a whole, accountable where the potential beneficiaries of employment schemes can organize and speak out (rule 10), and, in general, where there is a functioning, free and noisy "civil society". This is enjoyed by Maharashtra, and by many other places with public works programmes against poverty, such as India as a whole, Botswana and Chile. However, in less open polities that attempt employment-oriented public works, an accessible review process is needed even more; and the prima facie case against public monopoly of works execution and management is stronger.

Apart from the execution of the public works themselves, the nature of the nearby labour market is important to the secondary effects. The effect on the poor of decisions about public works wage rates, timing, location, and much else, depends in large part on the variability, price elasticity and competitive structure of supply and demand. This applies especially to the categories of labour demanded by the public works, but also to types of labour, skill, and (perhaps most important) capital equipment that might displace or complement this labour. Specific implications of these issues are discussed elsewhere in these sections. Here, we make five general points.

· The responses in private labour markets, to changes in the seasonal availability of labour due to public works, can hugely affect - double or destroy - the poverty impact of public works.

· Therefore, the impact on employment, poverty and output of such responses must be estimated (roughly, but with analysis of sensitivity to error) before starting work on major public works.

· Scheme planners should assume that private firms and households, in search of income and security, will respond intelligently to new wage, employment, and other options and incentives created by public policy.

· In economic terms, such responses will involve changes in output, as well as in labour supplies and demands, that reflect changing scarcities (and therefore prices) and risks, overall and seasonally.

· There may well be a political response; for example, big farm employers may well find it worthwhile to lobby for works that improve their irrigation, will oppose works that increase peak-season labour scarcity and wage rates, and may not feel strongly about works that improve small farmers' irrigation.⁴⁶

46. Unless such works divert public funds from work on irrigation, etc., for larger farms - or, in the medium term, by creating new opportunities for deficit farmers on their own farms, lead to labour shortages on big farms.