There must be an optimal term for any government though it won't be the same everywhere. A government that is in power for a very short time lacks continuity or confidence to implement agendas; one that is in power too long becomes complacent.
Change, or the prospect of change, must come often enough to prevent either situation occurring. A fine balance must operate for effective governance.
The US works in four-year periods, British-style democracies like India in five-year periods. African republics generally have one-man, one-vote, once in a lifetime systems though South Africa appears an exception to that rule.
In India, the first and last year of an electoral term seem most important in terms of economic agenda. In the first year of its term, a government is psychologically most comfortable about taking unpleasant measures. In the last year, it is desperate to raise money to fight the next election. In-between, things drift.
That really seems to have been the pattern in the 12th and 13th Lok Sabhas at least. The caretaker NDA government, which operated after the 12th Lok Sabha dissolved, wrote a lot of legislation in 1999 pertaining to key areas like telecom, insurance and IT. It duly implemented these upon coming to power in November 1999 in the 13th LS.
But the NDA government next legislated and passed new telecom and power policies in 2003 when dissolution was already looming. It also went into overdrive on road-building and divestment fronts. The period in-between (2000-2002) saw little action despite droughts, UTI, KP and Tehelka scams and the steady erosion of NDA support in Vidhan Sabhas.
Would Indian governments become more proactive if electoral terms were reduced to two or three years? One obvious danger: In many cases, the beneficial effects of new initiatives are only visible in the long-term. Governments with short electoral terms would ignore these possibilities and concentrate on stuff that pays off quicker in terms of moolah and votes.
An extreme example: the demographic timebomb embedded in Europe's social security. The pension benefits of a growing population of retirees are funded by progressively smaller workforces. The danger has been obvious for decades. But governments across Europe continue to ignore it since it will only become a problem after their respective terms end.
Urban infrastructure such as water-supply improvement and new metro rapid-transit systems can take decades to pay off. Fiscal and tax-reform like VAT also result in higher efficiencies only after several years. With shorter terms, there's no chance of such agendas being implemented.
Anyhow, it would require an Act of Parliament to change periodicity and one can't see sitting members voting to shorten their periods of privilege! But it does seem likely the 14th LS will start with a shaky coalition if one can believe the exit polls.
In behaviourial terms, any government with less than say, 290 seats may assume its days are numbered. This could make it frantic to raise money to fight for the 15th LS elections. That may be a good thing -- the easiest way of filling pork-barrels is mega-infrastucture projects and further PSU sell-offs. India needs these.