The flipside of corruption: State comparisons in India using public service delivery measures Amrita Dhillon March 31, 2020 No Comments Originally published on GI-ACE Understanding where corruption is high and why is a difficult task mainly because of the difficulties in measuring corruption. Given the well-documented problems with perceptions of corruption, recent literature has focused on objective measures of corruption. Focusing on the delivery of public services, a reliable measure could be the irregularities detected by top-down audits or social audits detected in such delivery. Unfortunately, as discovered by other scholars in the GI-ACE research programme (David-Barret and Fazekas, 2020), data on irregularities in delivery (“leakages”) or in procurement practices are notoriously hard to obtain, nor is it clear that data are correctly coded. However, digitization at all levels of government in India is proceeding rapidly so that perhaps there will be a rich source on audit findings in a few years. (Read more: India’s federal procurement data infrastructure.) In the meantime, we started our work on the premise that corruption is lower when the measures of delivery are good.1 There is a mass of publicly available data on the delivery of the two main programmes we are looking at: the NREGA (Mahatma Gandhi National Rural Employment Gurantee Act – a large workfare programme) and PMGSY (Pradhan Mantri Gram Sadak Yojana – a large rural roads connectivity programme). As a first step, we use this publicly available data to rank various states in India on how good their delivery was over the period from 2011-12 to 2018-19. The idea behind such comparisons is to highlight the determinants of good public service delivery, which is possibly linked to higher accountability. In this blog, I will discuss the first of these programmes: NREGA. The NREGA mandates the provision of 100 days of manual work on publicly funded projects to rural households in India. The Act envisions a rights-based approach: rural adults can demand work at a mandated minimum wage. The programme was initially implemented in the country’s poorest 200 districts in February 2006, with 130 additional districts added in the next stage (2007), and national coverage thereafter (2008). As of 2011-12, the Act provided employment to almost 40 million households at an annual expenditure of more than $8 billion, making it one of the most ambitious poverty-alleviation programmes in India to date. While the primary objective of the programme is social protection through the provision of employment, it also aims to create durable assets for the community, as a whole, and for socio-economically disadvantaged individuals (e.g., irrigation canals, ponds for water conservation, development of land for cultivation by socially disadvantaged groups, and other rural infrastructure). In order to fund the programme, the central government disburses funds as an advance, which is then backed up by receipts from the local authority. Implementation is carried out by grassroots institutions at the directly elected village, sub-district, and district levels under India’s decentralized system of governance. It is, however, difficult for the central government to get a full accounting of what has been spent and on what. There are a number of layers through which the money passes, leading to substantial “leakages” before reaching the intended beneficiaries (Banerjee, et al, 2019; Renikka and Svensson, 2004). Although official estimates of such leakages are missing, a survey in 2007-08 of 3,000 households in two large states of India showed estimates of corruption of the order of 75-80 percent (Neihaus and Sukhtankar, 2013). Although the success of NREGA can be measured against both the demand for work / wages and the quality of assets created, we cannot find reliable official statistics on the quality of assets so far. We, therefore, restrict attention to the success in fulfilling the demand for work and the wages offered both in terms of coverage and intensity. Our main variables are: Demographic coverage: total persons worked under NREGA as a fraction of the 2011 rural population below the poverty line; Financial coverage as the inflation-adjusted average yearly labour expenditure per person below the poverty line in 2011; Demographic intensity: the average number of days worked by participants; and Financial intensity as the inflation-adjusted average annual payment per participant. In addition, we include demographic and financial composite indicators that measure the state’s success in delivering both demographic and financial coverage and intensity. These indicator variables are aggregated at the state level. Since it is pointless to compare these variables in absolute terms without addressing the huge heterogeneity across states in demand for NREGA employment, we control for the demand for work with the proportion of rural adult poor who are in principle the main users of the programme. Finally, we also carry out the analysis for the restricted sample of those workers who belong to the disadvantaged castes as prescribed by the constitution. Looking at the average picture over the eight years of our analysis, our main results are as follows: On average, the demographic coverage is poor: most states cover only 1 percent of the rural poor signing up for NREGA. The best performing states are the northeastern states (Mizoram, Tripura, Sikkim). These states happen to be very small, so a more representative analysis of bigger states (population more than 25 million) suggests that Rajasthan, West Bengal, Madhya Pradesh, Andhra Pradesh, and Tamil Nadu perform the best, while Bihar, Goa, Punjab, Haryana, and Maharashtra make up the bottom five. Significantly, Bihar is among the worst performing states on coverage but performs well on intensity, while the situation is the opposite with Madhya Pradesh. On average, however, the correlation coefficient is high, with states that perform well on one indicator also performing well on the other indicator. Restricting the analysis to disadvantaged populations, the average performance is much better, showing that all states are targeting delivery to these groups. Figure 1. Summary of Indian states’ positions on the two composite indicators (coverage and intensity). Moving to a dynamic analysis, there are no huge changes in the state rankings over the eight years of the study. However, in intensity, West Bengal saw a spectacular improvement in its ranking from among the bottom ten in the first three years to the top five toward the end. More interesting is the question of what can be deduced from these broad patterns. States that are highly ranked happen to be those that are known to have high state capacity, considered to have three main dimensions: extractive (raising revenue); coercive (maintaining order and enforcing policies); and administrative (producing and delivering public goods and services) (Hanson & Sigman, 2019). Of the three dimensions, extractive capacity is negatively correlated with state performance, likely due to the fact that it is precisely the poorer (fiscal capacity) states who get the largest share of the NREGA allocation. Administrative and coercive capacity – measured by the number of police per capita, the rural literacy rate, and rural infant mortality rates – is more important to the success of NREGA, driven by the intensity variables. Our conclusions are very preliminary and should be interpreted more as suggestive evidence, being based on a very small sample of 19 states large enough to merit analysis. We plan to extend this work to a more disaggregated analysis, which may provide more insight.  Note however an important caveat: scholars have used differences between households’ level surveys and official statistics as a measure of leakages, most recently Imbert and Papp (2011). Official statistics may therefore not give a fully accurate picture. Amrita Dhillon Professor of Political Economy, King’s College London Amrita Dhillon has organized a number of workshops on topics ranging from Sovereign debt, reputational models in economics to a recent workshop on governance which was sponsored by the Journal of Public Economic Theory. Her training is in theoretical modelling, including political economy, public economics, game theory, and development. Dhillon received her Ph.D. from the State University of New York at Stony Brook; her main field of research is political economy. Recent work on corruption includes work on how electoral competition affects leakages in NREGA at the village level (Afridi et al 2019) and how natural resources can drive lower welfare via a political channel when compared to the right counterfactual (Dhillon et al, 2019). Leave a Reply Cancel replyYour email address will not be published. Required fields are marked *Comment You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong> Name * Email * Website Related blog posts Africa Integrity Indicators (AII) Team, April 15, 2020 Africa Integrity Indicators: 2020 Data Ready For Feedback Jacqueline Helen Harvey, November 20, 2019 Statistical challenges and ‘cash for penguins’ Africa Integrity Indicators (AII) Team, July 6, 2020 Africa Integrity Indicators: New data, a new round, and transitioning to Africa!