Commenting on Drèze’s DUET proposal, Martin Ravallion suggests that three steps are needed: learning from other countries’ experiences with similar policies, conducting trials in various contexts before considering scaling-up, and locating the case for intervention in a broader understanding of the problem of urban unemployment.

Many developing countries, including India, have a problem of urban unemployment. This is not new, though the pandemic and (in some respects) the policy responses have made things worse (in many respects). In the hope of addressing the problem of urban unemployment in India, Jean Drèze’s DUET proposal is a wage subsidy scheme to promote informal-sector employment in urban areas. Eligible public institutions receive ‘job stamps’ which they can use to pay workers, with the government covering the wage bill up to the statutory minimum wage.

Urban public employment programmes have often seemed like an obvious solution to urban unemployment. Just get the government to hire more workers. A famous example was in Kenya in the mid-1960s, when the government implemented a policy of providing extra government jobs in Nairobi, along with incentives for more private sector employment. So too have wage subsidies been tried often. There are differences in policy design, though a common feature is that the government subsidises, partially or fully, the wages of eligible workers. It seems to be common to link the wage subsidy to extra training efforts, and Jean’s proposal suggests this option.  

In assessing the case for such policies, a key thing is to understand the likely behavioural responses, including so-called ‘second-round’ and ‘general equilibrium’ effects. These can undermine well-intentioned efforts to create new urban jobs. For example, extra urban jobs will also attract rural migrants, hopeful of getting some of the new jobs. Indeed, under certain conditions, the count of the urban unemployed will expand, not contract. (In development economics, this can happen in the textbook Harris-Todaro model.) That is not just a theoretical possibility; indeed, that is what happened in the case of the Kenyan policy in the 1960s (as described in Todaro 1969).

On wage subsidies, employers can respond by simply replacing existing workers (or planned hiring) by subsidised workers, so as to cut their overall wage bill. It could be argued that this would be less of a concern for the public agencies that are to provide placement in Jean’s proposal. Maybe, maybe not. I can imagine that the placement agencies would also be keen to cut their costs, possibly to support other activities. That concern would only increase when (as mentioned by Jean) the idea is scaled up to include private non-profits. (The fact that an employer is ‘non-profit’ can still mean that it is keen to keep its costs down by diverting its workers to something like DUET.) 

At one point, Jean recognises this possibility, when he says “The list of works should not be so broad as to displace existing jobs in public institutions.” It is not clear to me how the scheme would be able to determine what is on the list in a way that prevented displacement. I can agree that, as Jean says, there is work to do, such as in maintenance of public facilities. Are we to presume that the public institution is spending its money unwisely, underspending on maintenance? If so, why would the public institution not simply shift its current (limited) maintenance efforts to DUET workers, leaving money for that (unwise) stuff it prefers to spend on. Or is the problem just too little money in the aggregate going to the local authorities? Then why not simply spend the (potentially considerable) sum required to finance DUET on more budget support for local authorities?

I would suggest that three steps are needed if this is to be considered seriously. First, it would be a good idea to look into the literature for clues as to what can go wrong in past efforts of this sort, and it is clear that things can go wrong. While I have not reviewed this literature for some time, I expect that India can learn from the experiences of other countries with such proposals. There have been evaluations of public employment and wage subsidy/training programmes in various countries, though my impression is that the record is rather mixed. As mentioned by Jean, it is hoped that DUET will morph into an urban employment guarantee scheme. Here too I would tread carefully, especially given the rather mixed evidence on India’s famous Rural Employment Guarantee Scheme (MNREGA), which has worked better in some places than others, and (alas) often less well in poorer rural areas. And if MNREGA has been as successful as Jean seems to think one has to wonder why there is so much unemployment of relatively unskilled (informal-sector) labour in India’s urban areas.

Second, trials should be done in various contexts before considering scaling up. Jean mentions that “DUET could be easily initiated on a trial basis in a particular district or municipality”. OK, but it is crucial that we can reliably learn from such trials. I certainly don’t think that randomised controlled trials (RCTs) are the ‘gold standard’ that some influential advocates have claimed, but that is a tool that could be well-suited to the evaluation of DUET trials. As an example, one might check out the RCT reported in Galasso et al. (2004) of a wage subsidy and training scheme in Argentina. The study found a small, but statistically significant effect on employment, but not incomes. The training component had low average impact, though somewhat higher for workers with prior training. Importantly, qualitative work (as a complement to the RCT) revealed that the wage subsidy scheme worked very differently in practice to expectations.

Third, and importantly, one must put this idea in a broader evaluative context. What I find missing from Jean’s proposal is even a hint of an analysis of why the problem exists in the first place. This must involve a deeper understanding of what is holding back employment in the private sector, relative to the supply of labour to that sector, which also depends on what has been happening in the rural sector. If we simply observe the existence of urban unemployment, something like DUET may seem the obvious solution. However, as I have emphasised, that is not so obvious. There is the Harris-Todaro aspect, which points to the role of minimum wages and other restrictions on urban labour markets, and also warns us that extra government jobs need not yield lower urban unemployment. I do not take the view that removing the so-called ‘distortions’ in urban labour markets is necessarily feasible or even desirable; that remains an open question. But I do think we need a broader discussion of why this problem exists in the first place, and whether extra public employment and training, along with a government subsidy, is the best solution, or would even work at all. 

Further Reading

• Todaro, Michael (1969), “A Model of Labor Migration and Urban Unemployment in Less Developed Countries,” American Economic Review, 59(1):138-148.
• Galasso, Emanuela, Martin Ravallion and Agustin Salvia (2004), “Assisting the Transition from Workfare to Work: A Randomized Experiment,” Industrial and Labor Relations Review, 57(5):128-142.