The 2019 Nobel Prize in economics for the proponents of Randomised Controlled Trials fulfils the discipline’s highest aspiration to mimic the laws and methods of natural sciences to explain social order. From the anatomical circular flow of income, the gravitational pull of rational behaviour, to the present application of experimental research for policy design, the discipline’s strong desire to foretell human behaviour and organisation with mathematical precision has progressed unabated. Experimental research designs are routinely used in clinical trials of new drugs and medical treatment to estimate the effectiveness of treatment. Such research designs are also used for impact evaluation of social and economic projects and policies. So, what is so novel about Banerjee, Duflo, and Kremer’s (BDK) method to deserve a Nobel? BDK propose the use of experimental designs not as an ex-post impact evaluation tool (efficiency study), but as an efficacy study, in a controlled environment, before a policy is rolled out – and hence the use of the term “controlled trials.” Thus, BDK gave a new spin to an established method, and started a “movement,” in Banerjee’s own words, of Abdul Latif Jameel Poverty Action Labs (J-PAL), where learned economists and social scientists experiment, evaluate, and counsel governments on what works and what does not, to uplift the conditions of the impoverished masses.

Several scholars have raised pertinent questions with regard to the method and ambitious claims of RCT and its application in policy design (see for instance, Reddy 2012, Basu 2014, Deaton and Cartwright 2017). My concerns relate to the philosophy of RCTs and how that transforms our understanding and experience of how policies should be made.

Policies have a life of their own. They germinate in the rich soils of people’s struggles and politics. They evolve, live, and become redundant when their purpose is served, or sometimes die a premature death. This process is the very essence of a functioning democracy. An ex-ante RCT breaks this organic process and places policy-experts at the epicentre of the policy design domain. The experts interact with the people, study their situation, engineer a policy intervention, and test the efficacy of the intervention. There may be several iterations of engineering and tweaking and testing, to improve on the efficacy before the policy is rolled out on a larger scale and take the shape of a ‘public policy.’ Such a clinical process is supposed to remove the unnecessary inefficient fluff that accompanies the somewhat chaotic generic process of making public policy and deliver a policy that really works, like clock-work, fresh out of the laboratory. There is no room for any agency of the masses, any role at all for them except as informants and as randomised participants of experiments.

This is not to say that experts should not or do not have a role in public policy design and process. Fiscal and monetary policies need domain expertise and are generally designed and implemented in a top-down approach. Similarly, large infrastructural projects are carried out by technical experts. However, in democratic societies, these policies also have to be sanctioned by the legislature as representatives of the people. It is important to understand that RCTs do not invade into these spaces, except in an ex-post scenario as a project evaluation tool. It would be impossible (for example, in an infrastructural project) or extremely costly (for example if we want to pilot-test two types of tax rules) to use RCT to test the efficacy of competing policies and suggest the best policy design in such cases.

RCTs have entered the broad arena of development policy, specifically anti-poverty programmes, to provide expert policy solutions where policies have historically evolved through people’s participation, long and gritty advocacy and political contingencies. What RCTs mean to do is to shortchange this lengthy political process and discourse and present a quick and unambiguous verdict. The entire policy process is depoliticised, with no representation of the people. What it also does in the process is, it gives tremendous power to a small group of policy experts and executives to decide the shape the policy should take. The process is undemocratic, unequal and elite.

In 2002, Banerjee along with Ghatak and Gertler had estimated the impact of tenancy reforms (operation Barga) on agricultural productivity in West Bengal. Operation Barga was a policy of the Left Front government in West Bengal where sharecroppers (known as bargadaars) were registered and were given heritable tenancy rights to land. They could not be evicted even if land was sold by the owner, land could not be sold without the permission of bargadaars and they were entitled to receive a part of the land price in such cases. The maximum rent that landowners could charge was also fixed at one-third of total output. By 2008, 1.51 million bargadaars were registered. Operation Barga, along with other land reform measures incentivised cultivators to invest in agriculture, particularly in groundwater irrigation and adoption of HYV seeds, and West Bengal achieved very high rates of growth in agricultural production (more than 6 per cent per annum) in the 1980s.

Imagine how a programme such as operation Barga could be part of a trial today. Knowledgeable (and well-meaning) economists would give different tenancy contracts to groups of farmers and estimate the impact on agricultural yields. The results would then be shown to governments and landowners, who would agree to change tenancy agreement, and peasants would rejoice! How different and wonderful history would have been.

In reality, however, the history of operation Barga goes back to the militant peasant struggles in Bengal – the Tebhaga movement in 1946. Land relations in Bengal were extremely unequal in colonial rule as a result of Permanent Settlement. Land was owned by Zamindars who owned entire districts together, and rack-renting was predominant. The actual cultivators or bargadaars bore the heavy burden of rent and exploitation of the zamindars and their sub-tenants. The Bengal famines in 1942-43 were the worst manifestation of the exploitative land relations in Bengal. Dissent among the peasantry against such exploitation led to the militant Tebhaga movement. Peasants demanded that agricultural produce be divided in three equal parts – two parts for tenants, one part for landlords. As a matter of fact, this was a recommendation made by the Floud Commission in 1940, but which was not taken forward by the colonial government. Tebhaga was the first state-wide movement organised by the Kisan Sabha in Bengal. Even after independence land relations did not improve significantly. Zamindari was legally abolished, but large landowners – the jotedars – retained land and high rents prevailed. The West Bengal Estates Acquisition Act was passed in 1953 and West Bengal Land Reforms Act was passed in 1954. Though the Act came into existence in 1953 and 1954, they were not implemented till the United Front government came to power in 1967. Growth in agricultural production in the State remained low and stagnant. The unequal land relations acted as major hindrance to the adoption of new technologies and agricultural investments. Politics in Bengal also reflected these antagonistic class relations. The Communist Party played a major role in organising peasants during Tebhaga, and the communist movement remained strong in the State, though state power remained with the Congress party dominated by big landlords. The 1960s and 1970s witnessed the emergence of radical left politics with the Naxalite movement. The tumultuous politics of the 1960s and 1970s led to the formation of United Front government in 1967 and finally the Left Front government in 1977. The major promise made by the Left Front was to undertake extensive land reforms, which it successfully carried out after it assumed office. Land reforms in West Bengal had three major components: imposition of land ceilings and redistribution of ceiling surplus land, operation Barga, and distribution of homestead land. Operation Barga fulfilled the demands of the tenant farmers raised during the Tebhaga movement. A programme such as land reforms could not be carried out without devolving state power to the people. Peasants themselves played a very important role in verification of land records, identification of benami land, verification of bargas and selection of beneficiaries. This was a unique feature of land reforms in the State. Local governance through panchayats was strengthened in the Statethrough the panchayats, local power relations were reversed and the landlords lost political traction in local politics.

Land reforms shook agrarian West Bengal out of its inertia. Agricultural investments increased sharply, short duration high yielding irrigated varieties of summer paddy (boro crop) was adopted across the State, and agricultural production and productivity increased. Banerjee, Gertler, and Ghatak (2002) showed that Operation Barga had a positive effect on agricultural productivity. The paper used district-level panel data for the period 1979–87 to study the relationship between district-level variations in the level of registration of bargadaars and agricultural productivity, while controlling for other factors, and found that “the tenancy reform program called Operation Barga explains around 28 percent of the subsequent growth of agricultural productivity there.”

A policy such as Operation Barga and the broader land reforms programme was the culmination of decades of peasant struggles. It was a ‘movement’ in the true sense of the term, a movement where peasants redefined Bengal’s political, social and economic relations. Had operation Barga not raised agricultural productivity, would it mean that the struggles of the people ‘did not work,’ even when it changed the power relations in the countryside and the political consciousness of the cultivating tenants forever?

References

Bakshi, Aparajita (2014), Land Reform and Access to Land among Dalit Households in West Bengal, in V. K. Ramachandran and Madhura Swaminathan (eds.), Dalit Households in Village Economies, Tulika Books, New Delhi.

Basu, Kaushik (2014), “Randomisation, Causality and the Role of Reasoned Intuition,” Oxford Development Studies, vol. 42, no. 4, pp. 455–72.

Deaton, Angus, and Cartwright, Nancy (2017), “Understanding and Misunderstanding Randomised Controlled Trials,” Working paper no 22595, National Bureau of Economic Research, available at http://www.nber.org/papers/w22595, viewed on November 4, 2019.

Ghatak, Maithresh, Gertler, Paul, and Banerjee, Abhijit (2002), “Empowerment and Efficiency: Tenancy Reforms in West Bengal,” Journal of Political Economy, vol. 110, no. 2.

Reddy, Sanjay (2012), “Randomise This! On Poor Economics,” Review of Agrarian Studies, vol. 2, no. 2.