However, given the limited geographical spread of such direct links to modern industry and commerce, in most areas the pivotal role in the expansion of rural non-agricultural employment appears to have been played by the expansion of government expenditure.
 
As noted earlier, the eighties were a period when, along with a rapid increase in all sorts of subsidies and transfers to households from government, there was a very large increase in expenditure on the rural sector by State and Central governments, and this was also a period when the expenditure on Rural Development expanded manifold. More generally, throughout the period political developments tended to give rural interests greater power and they were able to command an improvement in the historically low share of government expenditure benefiting rural areas. Although this improvement in share should not be exaggerated, an indication may be the fact that nearly 60% of all new government jobs created accrued to rural areas during the decade.
 
Moreover, NSS data suggest that, despite a low average contribution of only around 5% of total rural employment, the government's contribution was around a fifth when it came to either total rural non-agricultural employment in 1987-88 or the increments in total rural employment between 1977-78 and 1987-88. Moreover, in 1987-88, about 60% of the regular non-agricultural employees in rural areas were employed by the government which created almost 80% of the increments in such regular jobs during the decade covered.
 
Thus, the total quantum of increased flow of public resources into rural areas must have been significant. This flow of resources took two predominant forms. There was, first, a fairly large expansion of `rural development' schemes with an explicit redistributive concern. This included not only the various rural employment and IRDP programmes but also a plethora of special schemes for a variety of identifiable `target' groups. Motivated by the realisation that income growth by itself would not `trickle down' in adequate amounts, these programmes were however less than entirely successful. They spawned a large bureaucracy and they became a focal point for the politics of `distributive coalitions'. Yet, even though the intended beneficiaries often got short-changed because of such leakages, these programmes represented a fairly massive net transfer to rural areas.
 
The second avenue by which resources flowed from government to rural areas was through the greater accessibility of the rural elites to the varied benefits of government expenditure. In part this was a result of greater mobility due to better transport infrastructure, but to a large extent it was also the outcome of the fact that with governments changing frequently (particularly at the state level) more new favours, not just jobs, but also various types of agencies and contracts, had to be distributed more often and the rural areas got a greater than normal share in such largesse. The resulting flow of resources and the consequent generation of rural demand led to growing opportunities for diversification of the self-employed from agriculture to non-agriculture.
 
To a very large extent, the direct access to government permanent employment and also to many of the other resources was confined to the better-off and more powerful groups in rural society, to whom such incomes were more lucrative than agriculture. It should also be noted that such access to better employment or other resources was dominantly accorded to male workers rather than to women workers.
 
Associated with this was a large and significant increase in the proportion of the 15 to 29 age cohort which continued in education rather than join the work force. Once again, males dominantly benefited, but the increase is evident even in the case of females, albeit to a much smaller extent. In part this shift to higher education rather than seeking immediate employment must have been a result of the expansion of educational facilities as part of the general expansion of government in rural and semi-urban areas. But this may have represented also a motivational change (to acquire necessary qualifications for a regular non-agricultural job) among the youth - particularly the male youth - in the relatively well-off sections of rural society.
 
There was thus a movement out of agricultural work at the margin by workers and potential workers from such better-off rural groups, which meant that sections of the relatively rich vacated agriculture either to obtain regular employment, mainly in the service sector, or to take up non-agricultural self-employment. This increased the ability of members of the less well-off rural households - including and in some areas especially, women - to find agricultural work, and also created a demand for certain types of rural services and industry.