The Centre-State Political Transfer Cycles

Using Arellano-Bond dynamic panel-data estimation methods (GMM) for a balanced panel data from 1980-2010 for 14 Indian states, we try to find whether the election affects the individual components of transfers from the centre to the states, namely, grants from the centre, loan from the centre, and tax devolution. We also attempt to examine, if different transfer variables and other politico-economic characteristics of the country are able to create the possibility of retaining the political power for the incumbent. We find that the generally right wing and coalition governments are relatively less likely to transfer resources to the states. However, the state level ruling party, which is either the same party at the centre or an ally, tends to get higher transfers from the centre than a non-allied one. Similar to the pre-election political budget cycles found in the existing literature (Drazen and Eslava, 2010, Aidt, Veiga and Veiga, 2011, Klomp and Hahn, 2013, Chortareas, Logothetis and Papandreou, 2016), the political transfer cycle (PTC) is visible one year before the parliamentary election, whereas cycles are visible in the year of assembly elections in case of grants from the centre and loan from the centre. The tax devolution does not show any clear pattern of cycles, either in parliamentary or in assembly elections. The paper is also extended to a binary Logit specification to test for the incumbent’s probability of winning the election. We find that grants and loans are likely to have varied impacts on an election win depending on the timing. Opportunistic grants in the year before the election are likely to help in winning whereas; a loan punishes the incumbent in parliamentary elections. Instead, opportunistic manipulations in grants and loan in the year of election help the incumbent retain political power in national elections whereas, only the opportunistic loan from the centre to the states in the year of election help to win the assembly elections. Further, it is found that a higher voters’ turnout in the state is more likely to win the election, inflation reduces the possibility of the win, and a more experienced government has a higher probability of a win. Moreover, our results also show that the right wing government is more likely to win the election. Length: 51 pages