When does a country reform?

A serious account of reform must explain three things:

Why bad policies persist.

Why reform is resisted even when persistence is costly.


What tips a country from one equilibrium to another at a particular moment.

This framework comes from three papers published in the early 1990s. In 'Why Are Stabilisations Delayed?' Alberto Alesina and Allan Drazen show that a country on an unsustainable path can postpone reform. The obstacle is not disagreement over whether action is necessary, but conflict over who should bear its costs.

Each organised group holds out, hoping a rival will concede first. The deadlock ends only when, for one pivotal group, cost of waiting another day exceeds the cost of giving in. This is a war of attrition.

Beneath it lies a subtler trap. In 'Resistance to Reform: Status Quo Bias in the Presence of Individual-Specific Uncertainty', Raquel Fernandez and Dani Rodrik showed that a reform a majority would support after it's brought in, can be rejected before it's implemented if individuals can't know in advance whether they will emerge as winners or losers. Reforms that identify their beneficiaries, and credibly compensate their losers, travel further than those that promise aggregate gains.

What usually dissolves the deadlock is crisis - a policy failure made too expensive to ignore. India has rarely moved without one. So, a country reforms when:

Cost of status quo rises rapidly and falls on someone with the power to act.

A decisive actor expects to be better off after reform than from further delay.

Uncertainty over winners and losers is low enough - or compensation credible enough - that fear of losing no longer blocks change.

Political horizons are long enough to realise gains that reform almost always back-loads.

Remove any one of these conditions and the window opens onto a wall. This is why no single factor is sufficient, crisis least of all. In 'Do Crises Induce Reform?' Allan Drazen and William Easterly tested the hypothesis in more than 150 countries. They found that high inflation and black-market premia are often followed by reform, but large fiscal deficits, current account gaps and negative growth are not.

Crisis is neither necessary nor sufficient. It works only when it raises the visible, fast-moving cost of inaction beyond one group's concession point. Slow crises can persist indefinitely.

Set this framework beside the present. In 2025, GST was compressed to 2 rates, a new I-T Act was introduced, 4 labour codes were revived, and a sweeping deregulation drive gathered pace. None responded to a domestic crisis. They answered external pressure - US tariffs that briefly reached 50%, a war premium on crude, and global search for manufacturing alternatives to China. The world raised the price of Indian inertia, and the logjam broke.

Which sharpens the worry. There is now an interim tariff understanding with the US, crude prices have eased, and the West Asian ceasefire largely holds. The one force that has most reliably driven Indian reform - crisis - is receding on every front.

Drazen and Easterly found that the appetite for reform fades as pressure that produced it subsides. Reforms yet to be attempted are the hardest ones, with losses concentrated in land and labour, and many lying on state and concurrent lists.

If crisis can't be relied upon, the alternative is to manufacture commitment. Finn Kydland and Edward Prescott showed that optimal policy is time-inconsistent. A government free to change its mind tomorrow cannot make a credible promise today. The solution is to tie its own hands.

India's most durable reforms have done exactly that: the 4% inflation target written into the RBI Act in 2016; IBC of the same year, which put liquidation beyond the defaulting promoter's reach; and GST Council under Article 279A, through which states pooled enough sovereignty that none could defect alone. The deepest reform is not a policy but an institution that removes the discretion to reverse it.