I have to add some reflections to Trivellato's extremely clear words.

First of all, I must emphasise the clarity of his address. I have followed carefully most of the meetings of the Study committee on the Appraisal of the effects of public policies.

Today I see that Trivellato has set out the problems and methods with absolute clarity, even for those who are not experts in this line of work, since there is no attempt to hide - and this is the key point to stress - the complication and sophistication of the proposal that is being put forward. This is obviously a crucial point.

Having made this preliminary observation, I feel that as President of the CSS I am to an extent called upon to speak about the relevance of the work undertaken on the subject of evaluating the effects of "public policies", as per the meaning illustrated by Trivellato, with reference to the two basic aims of the Consiglio. Which are firstly to promote interdisciplinary work and the methods to be adopted with this sort of work, and secondly to help draw up and implement public policies, in other words to study and solve relevant and unresolved social problems.

Up until now, our "white papers" have been numerous over the past ten years, and always regarding critical issues. On the subject of appraisals, I recall - I suppose many of those present will know it - the white paper on the appraisal of research, the findings of a study committee coordinated by Alberto Zuliani.

These papers usually provide overviews of single issues and researches, but contain a number of recommendations to actors working in the area under review and to political actors working in the respective fields of analysis.

I believe that starting from the consideration that white papers are reports on the state of current issues, disciplines and researches, allowing the issue of appropriate recommendations to policymakers and operators in the field, working on the evaluation of public policies must mean expanding the "toolbox", to borrow from Wittgenstein, of the operators of practical life, in this case administrators and governors.

In this sense, I fully see the role of the Committee, which seeks to establish a precise method, while there is talk of evaluating public policies in various and multiple ways.

Trivellato's report is very clear and intellectually very honest: in specifying the exact field of application of this evaluation method, the characteristics that actions to which the impact evaluation refers should presumably have are highlighted. This raises the problem of the dimension of the target population, the sort of "treatment" that the individual area of intervention must have and the likely outcome to be ascertained. All of the above through a set of methodological reflections that appear to be and indeed are sophisticated and complete. The type of intervention tends towards the micro level, to the exclusion of the macro, while we are used to seeing at all times, on the television screen, indexes and figures regarding the macro economic situation, presented more or less correctly.

I believe that all the preconditions have to be satisfied if we want to evaluate the effects of the action.

This necessitates an adequate database, and the further it goes back in time the more effective and efficient it is for evaluation purposes. It is here that there arises the possibility of fully applying, and not just metaphorically, that "counterfactual" paradigm that lies at the heart of the presented method.

There emerges a clear impression that we are talking about a ‘frontier' contribution that sciences, let's call them statistical sciences, are offering to other scientific disciplines and to policies in order to tackle a problem that is absolutely real, true, heartfelt. Albeit in very general terms, as a rule.

The underlying sense is that, having explained the method, having explained the possibility of obtaining important answers using this method for the purposes of public policies - we remain at this general pronunciation, and it serves the purpose of understanding each other - the main thing is that having cleared up all this, those who are governing and administering will be asked to use these instruments, not directly, because the method is one for professionals, but to turn to the professionals of this method in order to finetune their ability to do politics, to administer.

I think this has to be pointed out after the many Committee discussions, and this is a point on which a number of different opinions have been voiced.

Having cleared this matter up, there is however the question of understanding the limits and why a refined instrument for evaluating effects is useful, and to whom, to which section of the ruling class and the political and administrative classes it is relevant.

Posing such a question also means, to an extent, seeking to understand why interest remains so modest. In short, although it is true that other countries have more experience, I would not say that these experiences are proof of a widespread interest.

As Ugo Trivellato said about France, they have indeed started (and we too hope to have started, and to have a follow-up to this work), but we must recognise that we are in good company, as regards the moderate interest shown in this very precise methodology.

One reason for this lack of interest undoubtedly lies in the difficulty, or sophistication, not just of the method described today, a cutting-edge method that has never before been proposed in our country - but of all the evaluation models and paradigms that require, for instance, a tradition of statistics within the PA. Such a tradition is lacking.

In my long experience of public administrations I could cite many examples to demonstrate that this culture does not exist. Even the reform of ISTAT, the work of Nino Andreatta - and I had the pleasure, many years ago, of following the process - did not produce a widespread statistical culture, and here I am not talking about the evaluation of effects, but merely of the premises, of the preconditions.

The real problem lies in the way the administrative political system works. This is the only part in which I disagree with Ugo Trivellato, since the final summary on defects contains a moralistic, yet very widespread, judgement, founded without explaining a lot, on a general condemnation of the flaws of the political classes and of institutional mechanisms.

I believe that, together with an effort to evaluate the effects of public policies in terms of the contribution that the scientific world of social sciences must make to the world of governance and administration, another equally lucid analysis of the reasons why relative interest is so modest is needed.

What does it mean to propose a serious appraisal of the effects of public policies?

It means starting with a view of social problems that are objectively complex, and which should thus be tackled using adequate, complex means.

Yet what is the response of politics, beyond the contingent response, one that is consolidated over time in our country and in others? The tendency to simplify. It is necessary to simplify political talk, now that the time of ideologies that in turn simplified political talk is over. Today the message is that of using, for political debating, the figure of the media debate, based on the logic of generalist television, which is particularly reductive. There is still no reflection, for instance, about the usefulness of debate and of interactivity through diversified multimedia tools.

The question of maximum simplification is increasingly topical, and there is a second reason, mentioned by a writer on politics, Geoff Mulgan, who recently acquired experience close to Tony Blair, and who in his latest book, which is largely a good read, said: "then there is the problem of believing and getting others to believe that one can control and govern phenomena that in practice often cannot be controlled or governed". In short, it is not admissible in politics to give speeches that are not in some way simplified.

Another point to examine is the reality of the parliamentary world. Back when I was a member of the Court of Auditors, and when I had preached its use, which was basically an original aim of the Court, as an operating arm of Parliament, and then having worked sometimes in Parliament as a sort of liaison officer with the Court of Auditors, I saw, just as I had seen when studying the problems of parliamentary control, that the job of a parliamentarian is one that necessarily focuses - this is the culture, and the expectation - on law making. The merit of the parliamentarian, in relations with his party, his constituency, his interest group, is that of helping to pass laws, even if it is only to add his vote.

This is also basically the attitude of governments, for the reason given by Mulgan, that of always giving the impression that something is being done, that in any case there is a law to deal with a problem, to have it under control. But you will never find a real interest, in programmes and in attitudes, in the administration, in travelling around from administration to administration, to really understand what they are up to.

In France it is possible to write what has been written in the changes to the Constitution because since the 1958 Constitution France's Parliament is relatively unimportant. This is not meant as a polemical observation, it is an observation written in the manuals of constitutional law. If it is so, then it is right to assign to Parliament a control function, as has been re-endorsed and re-stressed in France's constitutional reform of last July. The context however is to some extent what I was saying.

It is right to talk about a prevailing formalistic culture in the administration, although in our Commission discussions I recalled that the formalistic legal culture has become almost a mentality born of despair, in relation to the stress to which the legal culture is subjected, the variety of sources that have to be coordinated, the fact that since the scope of administrative justice has been broadened the administration is always subject to a transfer of powers between direct administration and the administration of the courts. In short, a whole series of reasons for making me say it is true that the real reason behind much inertia and the refusal to modernise instruments lies only in formalism, which in practice has been absent for some time.

This all takes me onto another plane of considerations. In recent days I have completed a long work on the subject of "Building the institutions of democracy", the title I gave to a long study on the thoughts of Adriano Olivetti (to whom I am very attached, having known him when I was young, and for many years I was in charge of the foundation bearing his name). In doing this work I had to go back through all the topics of democratic institutions and how they operate.

I have the impression that institutions of a given type, those currently in place, which are in any case those inherited from historical representative parliamentary democracy, remain subject to the criticism raised sixty years ago by Olivetti in his book "Ordine politico e comunità" (Political order and community), which I have recently re-read.

A question that we must ask ourselves is this: what type of democratic institution needs a hardened, learned and effective evaluation of public policies? This is a question that we must ask ourselves.

There are for instance the numerous attempts, which at this point are not only ideas and projects, but that have also been interesting experiences of deliberative democracy and other forms of democracy that link the territory to the technological tools of the network, experiences that need to be examined in some detail.

Yet today we cannot go into the question of evaluating public policies without having an idea of the institutional work for which these methods can have a meaning and must be pursued.

I believe that we must attempt to link the evaluation of public policies to the question of a possible, more than likely, federalism in this country.

Whichever way you look at it, there is in federalism a competitive element, or soul; this competition is dependent on the effects of concrete measures. These are questions to be tackled.

It was also said, back in the times that I have been looking over again, that one of the reasons for federalism, or for a certain type of effective regionalism, was to check certain actions, where they are controllable and appraisable, since these could be pilot projects to be transferred to other regional contexts.

I believe that an outlook such as this should be encouraged, together with a more in-depth analysis of the reasons why the appraisal is not liked or not easy to adopt. It is in this direction that the contribution to be made by the Consiglio per le Scienze Sociali can continue.