The study's principle messages may be summarised as follows:
It is important to have an electoral mandate for reform. This is one of the strongest findings to emerge from the study. Reform "by stealth" has severe limits, and major reforms for which governments have not previously sought public approval, tend to succeed only when they generate visible benefits very rapidly, which major structural reforms generally do not. Crises do often create opportunities for such "reform surprises", as governments are forced to reverse course in response to an emerging situation, but unless the pay-offs are very quick, the sustainability of such reforms tends to depend on the ability of reforming governments to make the case for the desirability of reforms on long-term, structural grounds. Reforms justified chiefly as responses to an immediate crisis may be difficult to sustain once the crisis has passed.
The importance of meaningful mandates makes effective communication all the more important. Major reforms have usually been accompanied by consistent co-ordinated efforts to persuade voters and stakeholders of the need for reform and, in particular, to communicate the costs of non-reform. Where, as is often the case, the costs of the status quo are opportunity costs, they tend to be politically "invisible", and the challenge is all the greater. Often, it is fairly clear who will pay the price for a reform - which firms are likely to come under pressure and which jobs may be at risk -whereas it is not at all obvious who is paying for the status quo: it is difficult to identify firms that have never entered the market, sectors that have not developed or the workers whom they would have employed.
This points to the need for policy design to be underpinned by solid research and analysis. An evidence-based and analytically sound case for reform serves both to improve the quality of policy and to enhance prospects for reform adoption. The impact of economic analysis also depends on the source: research presented by an authoritative, non-partisan institution that commands trust across the political spectrum appears to have a far greater impact. Building such institutions can take time, as their effectiveness depends greatly on their reputation, but in countries that have them, their prior analysis appears to have enhanced the prospects for reform in particular areas.
Partly for these reasons, successful structural reforms take time. The more successful reforms in the study generally took over two years to prepare and adopt - and this does not include the "pre-work" done in the many episodes in which problems and proposals had been debated and studied for years before the authorities set to work framing specific reforms. By contrast, many of the least successful reform attempts were undertaken in haste, often in response to immediate pressures: when it comes to policy reform, more haste can indeed make for less speed. Pension reforms, in particular, often have relatively long gestation times, involving a considerable amount of careful study and consultation. While governments should be ready to use political "windows of opportunity" when they open up, this may create problems if it leads to excessive haste.
The cohesion of the government is also critical. If the government undertaking a reform initiative is not united around the policy, it will send out mixed messages, and opponents will exploit its divisions; defeat is usually the result. The case studies suggest that cohesion matters more than such factors as the strengt or unity of opposition parties or the government's parliamentary strength.
Concertation may be helpful but is no substitute for government leadership. Reform progress may sometimes be facilitated by intensive tripartite discussions involving the government and the social partners in a highly formalised process. However, firmness of purpose on the part of the government also seems to be a critical element of success in such situations. A number of cases suggest that a "concertationist" approach is unlikely to succeed unless the government is in a position to reward co-operation by the social partners or can make a credible threat to proceed unilaterally if concertation fails. Where the government is too weak to lead or is unwilling to do so, the social partners have little incentive to make concessions. By contrast, corporatist arrangements can work well where the government is prepared to lead. Indeed, in cases where the social partners are involved in the implementation of reform, it may be useful to have them involved in reform design as well.
The condition of the policy regime to be reformed appears to matter a great deal: some policy set-ups are more "ripe" for reform than others. Successful reforms of established policy regimes often appear to have been preceded by the "erosion" of the status quo through smaller piece-meal reforms or reform attempts; where the existing arrangements are well institutionalised and popular and there appears to be no danger of imminent breakdown, reform is far more difficult. The question of "reform ripeness" concerns not only the seriousness of the problem to be addressed but also the extent of awareness of the costs of the status quo and of agreement on the need for change - it is a question of communication, as well as economic analysis, and is therefore to some extent endogenous to the reform process.
Successful reform requires persistence. A further important implication of the finding concerning reform ripeness is that blocked, reversed or very limited early reforms need not be seen as failures: they may play a role in undermining the status quo and setting the stage for a more successful attempt later on. All of the more successful pension and labour-market reforms examined in this study followed earlier setbacks and many less successful reform attempts in all three domains can now be seen to have helped set the stage for subsequent, sometimes far-reaching reform initiatives.
The case studies also provide further evidence in support of some of the major findings identified by the Department’s earlier econometric work, particularly with respect to the impact of crises and the importance of sound public finances. Finally, the case studies cast some doubt on the oft-repeated claim that voters tend to punish reforming governments: the likelihood of subsequent re-election was little different between the more and less successful reform episodes, and the re-election rates for all governments in the study was close to the average for all governments at all national elections in the ten countries covered during 1992-2008 inclusive. While the number of cases involved is too small to permit generalisation, it is also worth noting that almost all the governments that successfully adopted and implemented reforms for which they had prior electoral mandates subsequently went on to win re-election.