An excess of incentives and institutions in development

One of the problems with modern development discourse is its dissonance with both history and reality. One example of this is the idea of using incentives to generate development outcomes. So, if people don't vaccinate, then incentivise them with payment (or lentils). If you want children to do homework, incentivise them with payment. If people don't quit smoking, pay them to quit. And so on. In general, if you want to get people to do something, incentivise them with conditional cash transfers. 

The body of work on getting people to do stuff which advances development causes revolves almost completely around rewards/incentives and nudges. There is an asymmetricity in policy response to such problems in so far as in contrast to incentives, there is little by way of plain vanilla stuff like just making people abide by rules and enforcing those rules.

In this context, in a recent paper, Edward Glaeser points to how nineteenth century New York got people to connect to water and sewerage networks. He writes on the role of mandates and their enforcement in utilisation of infrastructure,

In 1842, New York City opened its Croton Aqueduct, which brought clean water into the city. This engineering achievement did not, however, end New York City’s Cholera problem. Somewhat surprisingly, the city would continue to suffer from cholera outbreaks for another 24 years. The persistence of cholera was not a puzzle to Dr. Stephen Smith, who led a team of doctors during the 1860s that trooped through the city and produced the 1865 “Report of the Council of Hygiene and Public Health of the Citizen's Association of New York Upon the Sanitary Conditions of the City.”

New York City had piped water, but water connections were expensive and tenement owners avoided the expense. Poorer renters lacked both resources and the incentive to internalize the wider health benefits of sanitation. New York City even had about 2,300 hydrants that dispensed free water, but using hydrants requires carrying water significant distances. Consequently, poorer New Yorkers continued to use shallow wells and pit latrines and they continued to die from cholera.

Smith’s report then produced the legislation that created New York City’s Board of Health, and Smith became its first leader. He began a system of inspections and fines that pushed tenement owners to connect to the water system. The Board’s inspection service was independent of the corrupt police force controlled by the Tammany Hall Machine. The doctors who set atop the Board’s leadership seemed to have been reputable men, like Smith, whose preferences (and reputations) kept them from striking corrupt bargains with those property owners who preferred not to pay for clean water.

This has resonance with all cities across developing countries. Even when water lines or underground sewerage systems are laid, very few households take the connections. The incentive-compatible approach that development economics advocates revolves around finance (and now nudges) - pay people to take connections, lotteries, lower connection charges, amortise connection charges with bills (pay-as-you-go) etc. 

I am inclined to argue that these are logically neat and theoretically robust. Besides they are amenable to guilt-free and elegant field experiments and research publications. In short, this is all fine for armchair theorising. Nothing beyond that. 

Let me layout a conceptual framework on thinking about this issue. 

Social contracts are two-sided. Citizens and governments agree to a contract. Citizens demand certain public goods and services and pay taxes for it. Governments use the taxpayer money to produce/provision those public goods and services. There is often a gap, for whatever reasons (including poverty and access), between delivery of those goods/services and their utilisation or uptake. Even when access has been provided, utilisation often remains elusive. In such cases, citizens have to be made to utilise the good/service offered. 

This process of making citizens utilise what is offered is a combination of rules/regulations/mandates (or exercise of state capability) and incentives. As the example of water and sewerage in New York shows, on most development issues and from historical examples of developed countries, ensuring utilisation has been largely about rules/regulations/mandates and their enforcement. Incentives have mostly been marginal contributors, relevant mainly in mopping up the few holdouts. This holds just as much to vaccination or school enrolment, as it does to taking water/sewerage connections. 

In the case of most public goods/services, history points to rules/regulations (in the form of exercising state capability) as the preferred choice to ensure utilisation. Development economics, on the other hand, is blind to rules/regulations and is focused on rewards/incentives, and in recent times on overcoming human psychological biases (the nudge factor).  

In simple terms, on the issue of utilisation of public goods/services (and more) development economics skirts around the difficult to grasp and even more difficult to implement issue of state capability and settles down on the comforting and logically neat ideas of measurable and testable incentives and nudges. 

In many developing country cities, utility connections have less than 20% households coverage in streets with water and sewerage connections, especially with latter, even in the middle class colonies. In low income areas with public taps, water connections uptake is similarly very low (even when connection charges are low). Unlike a case where just 10-20% of households are holdouts, and would perhaps need some incentives or subsidies, a case where just 10-20% have taken connections demands enforcement of rules. 

This World Bank working paper is a good example of a study that lists out as reasons (see the table in page 12) for low coverage of water and sewerage connections everything except the simple fact that many people just don't want to change the status quo and take a connection. And they don't face significant cost in maintaining the status quo. It's a teachable exhibit - it uses the word 'incentives' 36 times and the words 'enforcement' and 'rules' just six times each! Its 128 pages is also a classic example of over analysis in the context of development problems. Incidentally, it's also replete with several examples from developing countries on successful experiments with increasing utilisation through the likes of incentives, but none from developed countries about how they achieved the same. 

Incidentally, this is a very rare experimental study on the effectiveness of disconnections in ensuring utility bill payments. 

This is an unfortunate distortion, a deep negative externality imposed by excessively theoretical and deeply disconnected discourse that dominates international development and aid agencies. It has crowded-out serious engagement with hard issues of state capability improvements and crowded-in massive aid spending into mostly marginal and even silly ideas tailored around financial incentives and nudges. 

None of the above is an advocacy for strict penalties and hard enforcement, and avoiding incentives. But the reality is that when the baseline of compliance is so low, even among those who can afford to comply, then the first line of engagement should be enforcement. We can get to incentives and subsidies when we come to those who clearly cannot afford.

The anecdote about New York is also important for another aspect. It highlights the importance of individual agency. The personal initiative and role of Dr Stephen Smith in bringing forth the legislation and also working to enforce its implementation was critical in New York's success in delivering clean water. 

In the 18th Brumaire, Marx pointed to the limitations of individual human beings and the importance of circumstances and legacies. Extending that logic, development theories focus excessively on institutions - organisations and systems - and somewhat less on contexts/circumstances and historical paths (path dependencies). But it more or less marginalises the role of individuals. 

There is very little systematic research on the importance of individual (mostly non-political) bureaucratic or technocratic leaders in transformational development. This is surprising. History is replete with names like Stephen Smith, Robert Moses, Joseph Bazalgette, Baron Haussmann etc whose contributions were immense. Or the numerous Latin American mayors in cities like Mexico City, Sao Paulo, Curitiba, Medellin, Bogota etc who left indelible marks on their cities. Each country has their heroes among bureaucrats and technocrats.

Closer home, India has its names - KL Rao, E Sreedharan etc. There are several IAS officers who have played remarkable roles in transforming their terrains. It's common place for institutions and districts/corporations to be meandering along for years, only to be galvanised by the arrival of a dynamic officer who undertakes far reaching reforms and implements them with militant discipline and commitment, and for the system to revert back to its initial somnolence once the same officer departs. There cannot be anymore definitive evidence of human agency. And India's development history is full of such instances across geographical jurisdictions, Departments of state and centre, and state and central public sector entities.