There is increasing acceptance in modern democracies that policymaking should be based on verifiable and robust evidence, rather than on past practice, anecdotal evidence or ideology. This development is a natural consequence of more open and transparent government.

Evidence should inform policy in all government decision-making. For example, the location and scale of new roads should reflect statistics and projections on current and future transportation patterns. The capacity of higher-education systems should reflect demographics and labour market demands. The redesign of the health systems should reflect measured need, quality and costs.

Robust evidence is particularly important in a time of tight budgets, as it can be easier to persuade disappointed groups to accept the outcome of difficult choices if they can see how evidence supports decisions.

However, taking an evidence-based approach to policy requires many changes.

First, those making decisions must be willing to seek out impartial evidence and analysis to inform their decisions. Specifically, they must avoid the temptation to gather only the facts that support decisions they wish to make, so-called policy-based evidence-making.

Second, marshalling the evidence requires good data. The past two decades has seen a great increase in the amount of Central Statistics Office data available on all aspects of the economy and society, facilitating robust and detailed analysis. The data comes increasingly from administrative records in government departments and agencies as well as from surveys.

False economies
Where budgets are tight, survey data must be collected in the most cost-effective way. But we must also avoid false economies. My generation recalls the error made in cancelling the 1976 census of population, only to have a census taken in 1979 to correct for this mistake. Let us hope that we learn from that mistake when decisions are taken on the 2016 census — as a country with a dynamic population and a fluid labour market, we need a census to provide accurate estimates of the population at local level so that we can plan for change.

Greater use is now being made of administrative data to inform policy. The potential of these large data sets is being explored through analytics — the discovery and communication of patterns in data. Identification of such patterns helps the public sector to design and deliver the most economically beneficial services.

In some areas, such as enterprise policy, we began doing this more than a decade ago. In other areas we have lagged well behind the rest of Eu rope. In many cases this is because Irish administrative data does not systematically include unique identifiers, that is, a code for each individual, business and building.

The creation of unique identifiers in Ireland has been discussed for more than two decades. Concerns have been expressed about data protection. However, other EU countries have demonstrated that these are surmountable.

Finally, we need to make greater use of economic analysis in decision-making. This need was recognised formally by the establishment of the Irish Government Economic and Evaluation Service in 2012. Its role is to build up key economic skill sets, and its creation reflects recognition by politicians and senior civil servants of the need for specialist and generalist skills in policymaking.

The service should raise the quality of economic evidence going into policymaking at all levels from concept to design, and from implementation to evaluation. It should also generate healthy debates among specialists, and between specialists and generalists. This in turn would help to embed evidence and analysis throughout government decision-making as part of public sector reform.

Frances Ruane is director of the Economic and Social Research Institute and a member of the council of the Statistical and Social Inquiry Society of Ireland, which is hosting a symposium on public policymaking at the Royal Irish Academy in Dublin at 6pm tomorrow.