The fickleness of popular choice has often been commented upon, but even so the verdict in Andhra Pradesh must make the political class wonder how someone like N Chandrababu Naidu, wired right down to the block level, failed to anticipate the popular disenchantment with, and rage at, his administration.

In November, when Naidu decided to seek a fresh mandate ahead of time, no doubt spurred by the hope that the sympathy factor following the assassination bid on him would carry the day, he must have thought he was on a sure wicket.

Six months later, those expectations have been proved to be wrong.

The question is, will the National Democratic Alliance, of which he was the mainstay, face the same fate on May 13?

What impact will the TDP's decimation -- any expectation that the people will have voted differently for the Lok Sabha seems misplaced given the extent of outrage -- have on the NDA?

The exit polls, while proving to be correct in the case of Andhra even if while erring on the side of caution, have put a question mark on the NDA's numbers as well. Will they also come true?

If they do, the NDA's woes on Thursday could easily mirror the TDP's on Tuesday. Its decision to go in for early polls was no doubt egged on by the favourable electoral verdict in December's assembly election in Rajasthan, Madhya Pradesh and Chhattisgarh, but five months down the line the Bharatiya Janata Party appears to face hurdles in its re-election.

If the verdict on May 13 goes along expected lines, the NDA, despite the bravado from its top guns, should fall short of a clear-cut majority. But it won't be curtains yet for the government: anything above 250 seats should see a new version of the alliance emerging, with legislative support from other political parties.

But the lesson from the voter -- that poor, neglected soul who comes into his own once every five years, and gets forgotten promptly -- will remain. Just as it remains for Naidu in Hyderabad. The voter's overwhelming message in 2004, it seems, is not against economic reform but who benefits from them. May 11, hopefully, will lead to a change in direction in the way the reforms have been carried out.

So what could a chief minister, or a prime minister for that matter -- who is the toast of the Western world, praised for his commitment to economic reforms, farsightedness, focus on development etc -- have done differently?

Since the election was fought on the plank of economic development, can a negative from the voter be construed as a vote against the reform process? Not really, since there is a broad consensus among the two main political streams in the country over the necessity of economic reforms. The voter in Andhra Pradesh, obviously, has chosen the Congress party's vision for reforms over the TDP's.

Barring 1996, every general election that has seen a change in government has had a strong impelling vote – 1977 (anti-Emergency), 1979 (anti-Janata Party), 1989 (anti-corruption), 1991 (anti-National Front), 1996 (anti-Congress), 1998 (anti-United Front). At least till it began, the latest round of election was characterised by the absence of strong public sentiment against the incumbent government.

The NDA has taken the absence of an anti wave as pro sentiment, and the outcome will be known in less than 24 hours.

But the implications for the NDA, particularly the BJP, if its margin were to slide, are enormous. For Prime Minister A B Vajpayee personally, a setback is the worst possible result. Through the six years of his rule, he has performed a delicate balancing act, tempering the extreme views from within his camp with the exigencies of running a coalition. No other BJP leader could have held this variegated quilt together -- and while it comes very easily to revile him for his various verbal pirouettes, his jugglery while not allowing the pace of administration to suffer needs to be acknowledged.

It is thanks to Vajpayee that the BJP has adopted development and economic reform as its main plank, such as it is conveyed by the India Shining campaign, over its more traditional, natural ones. Narendra Modi's 'experiment' in Gujarat has garnered enormous focus, the state likened to a 'Hindutva laboratory', but in all the noise and bluster of electioneering Vajpayee's own experiment has been ignored.

Modi's experiment has paid the BJP dividends in Gujarat, which explains the reluctance to remove him even in the face of the Supreme Court's strictures against his administration. If the party did not adopt his tried and tested formula -- despite the crazy fringe's insistence -- it was in large measure due to Vajpayee managing to convince everyone that his own vision --  development over division -- would deliver the results equally.

Given his unmatched popularity both within the BJP and outside, there was no way the prime minister's wish could be overruled. It was a Faustian bargain, and an adverse result for the NDA will mean Vajpayee's inability to keep his part of it.

Even if the BJP were to get more seats this time and yet be done in by its allies -- a more commonly accepted scenario for Thursday -- or even the fact that the NDA forms the next government despite its reduced numbers, will not make things any easier for the prime minister.

Put simply, a reduced mandate for the NDA will mean a weakening in the prime minister's bargaining capacity. Yes, the party will still need him, since there is no one else with greater acceptability within the BJP to lead a coalition. Nevertheless, it will mean an uneasy truce with the party, the hardliners merely biding their time.

Regardless of the results on Thursday, the prime minister will not have many options left in order to see his vision through. Whatever be the NDA's tally, he will have to form the next government. While a mere shortfall in numbers will only strengthen his critics within, ceding the reins of power to an alternative formation -- the worst possible scenario in this case: Sonia Gandhi as prime minister -- will mean the BJP turning its back on Vajpayee's path.