Corrupt States: Reforming Indian Public Services in the Digital Age Jennifer Bussell Lyndon B. Johnson School of Public Affairs University of Texas at Austin
Delivering Public Services Public service provision is a fundamental, but flawed, activity of government. Too often services fail poor people They are often inaccessible or prohibitively expensive. But even when accessible, they are often dysfunctional, extremely low in technical quality, and unresponsive to the needs of diverse clientele (World Bank, 2004).
The Perils of Public Services
Current Situation
Reforming Public Services Significant efforts to increase the quality of public services Two important trends: Privatization and public-private partnerships Increased use of information and communication technologies (computers, Internet, mobile phones) Goals: Improve accuracy, speed, and cost of services Limit bureaucrat-citizen interaction - Reduce corruption, especially speed money
A Global One-Stop Services Model Singapore ecitizen Centre Germany Buergerbuero (Citizen s Bureau) Finland Citizen Services One Stop Shop Brazil City and state centers, e.g. SAC Salvador South Africa Multipurpose Community Centres India State initiatives and Common Service Centres
Two Important Questions Why do (or don t) governments reform public services? Do public service reforms improve citizen services?
The Indian Puzzle Standalone, one-stop citizen service centers Computerized Frequent private sector participation Fundamental government services E.g. driving licenses, birth certificates, land titles, welfare benefits, tax payments Considerable variation across Indian states
State-level Variation Timing of Policy Adoption Sixteen states implemented during 1999-2006 Quantity of Services From 2 to more than 40 services Type of Services Socio-economic groups targeted - business licenses vs. welfare benefits Potential for corrupt rents - electricity bills vs. land records Degree of Automation Front-end computerization vs. full automation Ownership and Management Public, entrepreneur, and private company models
An Incentives-Based Approach: Direct Benefits Politicians faced mixed electoral incentives for and against reform Indian services are often cumbersome, inefficient, and corrupt $5 billion in bribes for services per year (TII 2005) Evidence that eservices can: Improve service characteristics (Bussell, 2009) Reduce corruption (Ibid.; IIMA, 2007) Clear potential benefit for the average citizen and anticipated benefit to ruling politicians
Corruption and Politics in India I: Indirect Costs Bureaucratic discretion and opaque processes enable siphoning of funds and bribe-taking in service delivery Politicians use transfer authority over bureaucrats to access bribes (Wade, 1985; de Zwart, 1996; Davis, 2004). Politicians use rents to fund future election campaigns (Ibid.; Sarin, 1999) More transparent services threaten bribes Extent of petty corruption affects the opportunity costs of implementing technology policies
Corruption and Politics in India II: Indirect Benefits Reform may also offer new opportunity for rents Private partnerships to run centers entail new contracting processes Large(r) scale bribes may be available from ownership and management negotiations Extent of grand corruption can affect the indirect benefits of implementing technology policies
Predictions Reform will require expected net benefits to ruling politicians Higher petty corruption states should implement policies later and in a less robust manner than states with less petty corruption. Higher grand corruption states should use private partners more often than states with less grand corruption.
Testing the Argument Controlled comparison of states Theory development - fieldwork in 7 states Andhra Pradesh, Gujarat, Karnataka, Kerala, Madhya Pradesh, Rajasthan, Tamil Nadu Out of Sample testing on 10 additional states Chhattisgarh, Delhi, Haryana, Himachal Pradesh, Maharashtra, Orissa, Punjab, Uttarakhand, Uttar Pradesh, West Bengal Interviews with government, NGOs, academics, and private sector Analysis of data from 20 states overall Bihar, Goa, Jharkhand Wide range of policy characteristics
Qualitative Findings Interviews in the states support the core hypotheses Overall: The threat of decreased income has been a problem (Central government official) Corruption and lack of transparency are barriers to egovernment (Private sector representative) Higher corruption: People do not want to shift to a new system because they have established ways by which they are able to skim money off the top (Haryana official) Lower corruption: We ve really had no problems from politicians the Chief Minister sees the benefits as greater than the costs (Chhattisgarh official)
Policy Timing Technology-Enabled Service Centers in India Cumulative State Policy Adoption
An Event History Model of Policy Timing Factors Influencing the Timing of Policy Initiation Variable Hazard Ratio z-ratio Corruption.37-3.06** Economic Development Human Development 1.86.88.01 -.88 IT Companies 1.03.18 Coalition.01-2.39* Electoral Competition 2.65 1.33
Quantity of Services
Corruption and Quantity of Services Independent Variable Model 1 Model 2 Model 3 Model 4 Model 5 Constant 24.7 25.8 26.5 32.7 22.6 Above Average Corruption Coalition -3.5 SDP per Capita Human Development Index Months since Policy Initiation (at end 2006) -14.0** -13.7** -14.0** 13.8** -13.29* (-3.22) (-3.12) (-3.16) (-3.45) (-2.43) (-.80) (-.65) r 2 adj.38.37.36.33.34-6.1 -.35 (-1.17.03 (.22) Entries are unstandardized regression coefficients with t-ratios in parentheses. *p <.05 **p <.001
The Consequences of Reform: Does it have any effects? Nemmadi initiative in Karnataka 800 privately-run centers Associated computerization of government offices Phased implementation Allows real-time comparison of traditional and reformed offices Stated goal: Provide transparent, speedy, and efficient services to rural citizens (Government of Karnataka, 2006)
Reformed Service Center
Reformed Service Center?
Citizen survey Evaluating Reforms 20 taluks, 1000 respondents, 3 types of centers Field experiment 18 taluks, 27 subjects, 3 types of services, computerized centers First comprehensive and independent evaluation Major goals: Unbundle the effects of computerization and privatization on service quality Maximize validity of causal inference while maintaining external validity
Service Characteristics Variable* Number of visits Total Minutes at Office Total Cost (INR) Days to Receive Service Computerized Taluk -2.78*** (-6.96) -53.17 (-1.96) -30.76 (-1.65) -20.66 (-1.82) Nemmadi -3.41*** (-8.76) -61.16* (-2.31) -55.22** (-3.05) -23.42* (-2.16) Age (yrs).01 (1.42) 1.48* (2.55).60 (1.50).28 (1.16) Education (yrs).00 (.14).64 (.38).22 (.19) -.84 (-1.14) Income (INR) -.01 (-1.71) -1.21*** (-3.69) -.07 (-.33) -.22 (-1.56) Urban.06 (.22) 111.58*** (5.64) 14.11 (1.04) 1.78 (.18) Constant 5.98 329.50 65.57 33.67 N 762 762 764 593 R 2.11.06.03.00 *Caste variable is included in model but not in table. Entries are unstandardized regression coefficients with t-ratios in parentheses.
Bribes and Efficiency Variable* Number of Officials Met Asked for Bribe? If asked, how much? (INR) Perceived Government Efficiency Computerized Taluk -.79*** (-4.77) -.00 (-.02) -68.87* (-2.37) -.47** (-2.76) Nemmadi -.85*** (-5.29) -.54** (-3.31) -102.98** (-3.53) -.34 (-2.06) Age (yrs).01*** (4.15).00 (.57).59 (.85).01*** (4.03) Education (yrs).02 (1.49) -.02 (-1.53) -2.21 (-1.09).02 (.13) Income (INR) -.00 (-.74) -.00 (-.91).14 (.32) -.00 (-.25) Urban.78*** (6.48) -.11 (-.89) 81.19** (3.13).36** (2.95) Constant 2.68 -.37 165.34 3.43 N 754 764 177 763 R 2.07.04.15.03 *Caste variable is included in model but not in table. Entries are unstandardized regression coefficients with t-ratios in parentheses. All Yes/No questions are scored 0-1, with Yes=1
Policy Implications Consider the institutional incentives underlying the established model Top politicians and street level officers Design policies to establish strong, if narrow, initial model Growing citizen demand affects electoral benefits Depend on design of policy, not partners, for success Evidence of corruption even in private centers
Research Implications Implications for understanding effects, and causes, of corruption Benefits of computerization & privatization require further attention Evidence and opportunities in country comparisons Cross-national Country cases
Thank You
A Measure of Petty Corruption State-level surveys conducted in 2005 by Transparency International India Sample of 14,405 respondents in 151 cities and 206 villages in 20 states Questions regarding experience with, and perception of, corruption (bribe paying) Largest-ever corruption survey when conducted; one of few to focus on petty/bureaucratic corruption
State Petty Corruption Scores (Scaled)
A Measure of Grand Corruption Audit of MPLADS spending 1997-2000 Inclusion of all grand corruption-style discrepancies in spending, e.g. spending of funds for inadmissible purposes or suspicion of fraud Calculation of per-constituency monetary value of inappropriate spending in each category Creation of overall index score, which was then scaled 0-1 for those states with technology-enabled service center policies
State Grand Corruption Scores (Scaled)
Ownership, Management & Grand Corruption Management Model Level of Corruption Below Average Corruption Above Average Corruption Public Model Haryana - edisha Himachal - Sugam Kerala - FRIENDS Rajasthan - Lok Punjab Mitra - Suwidha (1) (4) Entrepreneur Model Chhattisgarh - CHOiCE Kerala - Akshaya Orissa - esahayata Rajasthan - JanMitra W Bengal - Banglar Mukh (5) Uttar Pradesh - Lokvani Uttarakhand - Janadhar (2) Private Company Model Andhra Pradesh - eseva Andhra Pradesh - Rajiv (2) Delhi (CSB) Gujarat - Civic Centres Karnataka - BangaloreOne Karnataka - Nemmadi Maharashtra - Setu Tamil Nadu - RASI Uttar Pradesh - esuwidha (7)
State Corruption Scores (Scaled) Green = Grand corruption, Blue = Petty corruption