Contrary to what the states think, fiscal incentives are less important. In any case, with the introduction of the value added tax in April 2003, sales tax-based incentives will cease. However, the usual explanations may not provide the entire truth. This is clear from Punjab, which accounts for 0.7 per cent of approvals, or even Haryana at 1.25 per cent compared to 2.9 per cent in Orissa. Perhaps, the composition of FDI may provide part of the answer. In addition, since the initial base period FDI levels were low, a single proposal can distort the picture. For instance, Tamilnadu would not have been at 8.3 per cent had it not been for automobiles. On inter-state disparities, this is not specifically an FDI phenomenon, since domestic investment inflows and growth in state domestic product (SDP) also vary. However, SDP growth in many states in the 1990s is explained by agricultural rather than manufacturing growth and investment inflows are correlated with manufacturing and not agriculture. This explains why the shares of Andhra and West Bengal are relatively low. More pertinently, there is a band of backwardness and deprivation in central India, extending eastwards right up to the north-east. The hypothesis that voters will eventually vote out bad governance is long term, even if it has validity. Nor do labour migration or doles through finance commissions provide a satisfactory vent. If unrest is not to become endemic in the heart of India, schemes to diffuse socio-economic tensions will have to be found.