We end where we began: with the gap between what economics recommends and what politics permits. This module assembles the whole course into a practical question — how does good policy ever get adopted, and how can a reformer improve the odds? The answer is not better economics. It is better political-economy engineering of the reform itself.
Why beneficial reforms get blocked: status-quo bias
The collective-action module explained one barrier: concentrated losers out-organise diffuse winners. Raquel Fernandez and Dani Rodrik (1991) identified a subtler one that can block reforms even when a majority would gain. Suppose a reform will benefit 60% of people and harm 40%, but in advance no individual knows which group they will be in. If enough people fear they might be among the losers, a majority can oppose ex ante a reform that a majority would endorse ex post once the identities are revealed.
Fernandez-Rodrik: individual-specific uncertainty
Even a reform with net positive expected value and a majority of eventual winners can be rejected by majority vote beforehand, because uncertainty about who will win creates a bias toward the status quo. Identifiable certain losers will fight; potential winners who fear they might lose will hedge toward 'no'. This is why reforms that are popular in hindsight were unpassable in prospect — and why pilots, phase-ins, and information that resolves the uncertainty can unlock them.
Credible commitment and time-inconsistency
Kydland and Prescott (1977, the work that won the 2004 Nobel) identified time-inconsistency: a policy that is optimal to announce today may not be optimal to carry out tomorrow, and if people anticipate the reversal, the announcement fails immediately. A government promises investors it won't expropriate; once they have sunk their capital, the temptation to grab it is strong; investors foresee this and don't invest. The problem is not bad intentions but the absence of commitment. Reform that depends on a promise the government will later want to break is not credible and will not change behaviour.
The fixes are institutional: tie your own hands. Independent central banks, fiscal rules, international agreements (an IMF programme, a WTO commitment, an AfCFTA schedule), constitutional provisions, and delegation to rules-based agencies all work by making reversal costly — converting a breakable promise into a credible commitment. Much of institutional reform is, at bottom, the manufacture of credibility.
Compensation — and why it is hard
If reform produces net gains, the winners can in principle compensate the losers and everyone is better off (the Kaldor-Hicks logic of the cost-benefit course). In practice compensation is hard for political-economy reasons: the government must credibly commit to pay (time-inconsistency again — once the reform is done, why keep paying?); it must identify the true losers without inviting everyone to claim victim status; and the losers must believe the promise. When compensation works — visible, ring-fenced, delivered up front — it can convert organised opponents into supporters. When it is a vague promise, the losers rationally discount it and fight on.
Gradualism vs big-bang, and the role of crisis
- Big-bang — move fast and comprehensively while a reform window is open (a new government's honeymoon, a crisis). Advantage: exploits a brief political opening and makes reforms reinforce one another before opponents regroup. Risk: overload, mistakes, and a backlash that discredits the whole package.
- Gradualism — sequence reforms, build constituencies for the next step with the gains from the last, demonstrate benefits to resolve Fernandez-Rodrik uncertainty. Advantage: learning and coalition-building. Risk: opponents regroup; early steps get reversed before the payoff lands.
- Crisis-induces-reform — Drazen and Grilli and others argue that crises break the status-quo bias: when the old policy visibly fails, the cost of inaction rises above the cost of change, and previously-blocked reforms become possible. 'Never let a serious crisis go to waste' is a political-economy proposition, not just a quip — which is why so many durable reforms date to the depths of a crisis.
Ownership and the limits of conditionality
The structural-adjustment era (1980s–90s) is the great natural experiment in externally-imposed reform. The IMF and World Bank attached policy conditions to loans; compliance was uneven and reversals common. The lesson the institutions themselves eventually drew is a political-economy one: reforms imposed from outside, without a domestic coalition that owns them, tend to be reversed the moment the external pressure lifts, because the underlying interests never changed. Conditionality can support a reform that a domestic coalition wants; it cannot manufacture the coalition. Sustainable reform is built, not imported.
Exercise
Return to fuel-subsidy removal from module 1, now as a reform designer. A government must remove a costly, regressive fuel subsidy but fears the fate of governments toppled by exactly this move. Design a politically-feasible reform using the tools of this course. Address: (1) the concentrated-losers / diffuse-winners problem; (2) the Fernandez-Rodrik status-quo bias; (3) the credibility of any compensation you offer; (4) sequencing and the use of timing/crisis; and (5) why pure economic argument ('it's regressive and wasteful') is necessary but not sufficient.