The Political economy of MGNREGS spending in Andhra Pradesh

Are ostensibly demand-driven public programs less susceptible to political clientelism even when private goods are allocated? We investigate this conjecture using expenditure data at the local level from India's National Rural Employment Guarantee Scheme. By focusing on one state where accountability and transparency mechanisms have been employed and implementation efforts have been applauded, we do not find evidence of blatant vote buying before the 2009 election but do find that patronage played a small part in fund distribution after the 2009 election. Indeed most variation in expenditure is explained by the observed needs of potential benficiaries, as the scheme intended.