I posted a paper from NIPFP econs on the topic earlier. Here is another one this time from IGIDR econs - Arun Kaushik and Rupayan Pal. Both point to the importance of politics in allocating development expenditure. However, the results are just opposite.
The first one looks at whether fragmented or single party govts spend more (it is fragmented which spend more). This one looks at the role of political stronghold and does not think fragmentation has a major role:
This paper examines the effects of political factors on allocation of revenue budget for developmental expenditure by the sub-national governments, using data from 15 major states in India during the period 1971-2005.
It measures the ruling party’s political stronghold on the basis of constituency level electoral outcomes and shows that greater stronghold of the ruling party in a state leads to significantly higher proportion of revenue budget allocated for developmental expenditure. It also shows that voters’ turnout and political regime change have positive and significant effect on proportion of revenue budget allocated for developmental expenditure.
However, political ideology, within government fragmentation, disproportionality in representation, and effective number of political parties do not have any significant impact on budget allocation decisions of the Indian state governments. Results of this paper also indicate that greater reliance on market forces reduces the share of developmental expenditure. These are new and robust results.
The first paper looked at a time period of 1981-2005 and second one is from 1971-2005. NIPFP paper covers 14 states and IGIDR one covers 15 states (14 of NIPFP + J&K). So pretty similar states and time-period as well. Unless, we have a very different story in 1970s which changes the story so much.
The NIPFP paper stresses on coalitions leading to rise in expenditure. Its conclusion says:
The results of this study suggest that frequent formation of coalition governments is one of the major driving factors for increasing public spending of the Indian states in the recent decades. Ideology of a government also seems to be contributing significantly to this increase, but mainly through increasing current expenditure. Electoral cycle does not seem to be playing a significant role in influencing public expenditure. Probably in the era of competitive politics and coalition government, where governments are seen to be playing redistributive tactics throughout their tenure of governance, years just before election are no more treated differently. Well supported governments are found to be associated with increasing capital expenditure, whereas government spending on investment are found to be declining in the presence of a weaker opposition. In addition to these findings, results on fiscal space suggest that additional revenue only boosts public consumption, not public investment. These findings are robust to various forms of sensitivity analyses.
IGIDR paper stresses on single party consolidation leading to rise in expenditure:
We find that political stronghold of the ruling party significantly affects the budget allocation decision of the government. Greater stronghold of the ruling party in a state leads to significantly higher proportion of revenue budget allocated for developmental expenditure. In other words, preference of a government for developmental expenditure over non-developmental expenditure and grants-in-aid contribution together is stronger, if the major ruling party of that government has stronghold in larger percentage of electoral constituencies than that of other governments. It implies that the ruling parties with lower probability to be re-elected divert funds from developmental sectors to spend more on fiscal services, secretariat services, pensions for government employees, etc.
Moreover, we demonstrate that within government fragmentation and political ideology of the ruling party do not have any significant effect on allocation of revenue budget for developmental expenditure by the state governments. Effective number of political parties in a state and representativeness of the state government also do not appear to play any role as far as the state governments’ preferences for developmental expenditure is concerned. Our results are robust to alternative specifications of econometric model and estimation methodologies.
Which one to follow? NIPFP study looks more intuitive but that is just about it. Empirically, IGIDR one may be better.
I am not trained well-enough to spot differences in the methodologies. May be an enlightened reader can help.