Engagement with Economic and Social Research Institute

In the fifth meeting of the Select Committee on Arrangements for Budgetary Scrutiny we will meet representatives of three organisations. First, we will meet Professor Alan Barrett from the Economic and Social Research Institute. We will then meet Professor John McHale, chairman of the Irish Fiscal Advisory Council. He will be accompanied by Mr. Michael Tutty and Mr. Seamus Coffey who are members of the council. We will also meet Mr. Robert Chote of the UK Office for Budget Responsibility. He is in transit. This is the final public meeting of the committee. We can incorporate the issues arising from it and the replies due from Ministers into the series of recommendations that will form part of our report to the Sub-Committee on Dáil Reform.

The main part of our report will involve recommendations. The working paper circulated last week contained many of the recommendations to be considered. To avoid confusion, we will use the working paper as the basis of our work. Next Tuesday, 28 June, members will go through each recommendation and have an opportunity to add new recommendations, in addition to considering the document circulated. Further amendments arising from what we heard yesterday and what we will hear today will be circulated to members in advance of the meeting on Tuesday. Members should contact the clerk if they wish to introduce new recommendations. We plan on meeting at 4 p.m. and it is likely that we will need to have another meeting on Wednesday. It is also possible that we will need to meet again on Thursday in order that we can finalise the report. Is that agreed? Agreed.

I welcome the delegates, including those in the Visitors Gallery. I draw their attention to the fact that they are protected by absolute privilege in respect of their evidence to the committee. However, if they are directed by the Chairman to cease giving evidence on a particular matter and continue to so do, they are entitled thereafter only to qualified privilege in respect of their evidence. They are directed that only evidence connected with the subject matter of these proceedings is to be given and asked to respect the parliamentary practice to the effect that, where possible, they should not criticise or make charges against any person or an entity by name or in such a way as to make him, her or it identifiable.

I welcome Professor Alan Barrett from the Economic and Social Research Institute. He is joined by Mr. John Palmer from the Department of Finance and Mr. William Beausang and Ms Sarah Swaine from the Department of Public Expenditure and Reform. I invite Professor Barrett to make his opening statement.

Professor Alan Barrett

I thank the Chairman for inviting me to appear before the select committee and giving me an opportunity to discuss the task that has been assigned to the committee. The task of redesigning the routes and mechanisms through which the Oireachtas can provide for greater scrutiny of the budget has the potential to yield considerable benefits. As greater parliamentary scrutiny presents an opportunity for more evidence to be demanded in support of policy proposals, this, in turn, should lead to better policy outcomes. For this reason, I am happy to contribute to the committee's deliberations.

I am here in my capacity as director of the Economic and Social Research Institute, ESRI. Therefore, most of my remarks will relate to the ESRI's current role in the budgetary process and its possible roles in assisting the Oireachtas in the future. However, I have a direct personal interest in the ongoing process of redesigning the budgetary process, as I was one of the original members of the Irish Fiscal Advisory Council, on which I served from 2011 to 2015. I also chaired the first national economic dialogue in 2015 and will act as chairperson again next week when the second such event is convened. When I take questions from members of the committee after my opening statement, I will be happy to reflect on the possible lessons which can be taken from my experience in the Irish Fiscal Advisory Council and the national economic dialogue.

I would like to speak about the specific role of the ESRI in the budgetary process. According to its mission statement, its goal is "to produce economic and social research that informs public policymaking and civil society". In this way, a large proportion of its research has a broad relevance for budgetary scrutiny. For now, I will focus on the direct inputs which arise from two models that were developed by the institute and are maintained by us. I refer to the macroeconomic model which was formerly known as HERMES and is now known as COSMO and the tax-benefit model which is known as SWITCH. The SWITCH model which is used to assess the distributional impact of tax and social welfare changes is based on a large-scale representative sample of Irish households and allows us to calculate gains and losses across income distribution. It is critical that the model is based on a large-scale sample. In the absence of a tax-benefit model, analysts need to look at examples of households. However, it is impossible to know if the examples chosen capture large or small proportions of the population.

The SWITCH model is used by the Department of Social Protection and the Department of Finance, typically in collaboration with the ESRI, to simulate the distributional impact of tax and welfare changes. The ESRI tends to publish a distributional analysis of the budget in the days following the Budget Statement. The model has also been used in numerous research papers. The most recent example is a paper published last week which compared financial incentives to work in Ireland and the United Kingdom. I should note that the SWITCH model has the potential to be used in proofing the budget along a number of dimensions. I refer, for example, to poverty proofing, gender proofing and disability proofing. As the model is based on a large-scale representative sample, it is possible to draw conclusions that cannot be drawn from an analysis of example households.

I now propose to discuss our model of the macro-economy. We are in the process of retiring one model which was known as HERMES and introducing a new one which is known as COSMO. These macro-models have typically been used by the Department of Finance to conduct sensitivity analyses of budgets' central forecasts. Such analyses are published as part of the budget documentation. In budget 2016, for example, the ESRI simulated how the headline economic figures would change if global GDP growth turned out to be 1% lower than forecast or if the household savings rate turned out to be 1% higher than forecast. The Department of Finance also draws on the ESRI’s macro-model to simulate the effect of policy changes where the effect is large enough to be captured at the macro-level. For example, this year’s spring economic statement included a simulation of the effect of income tax reductions on employment. The macro-model has also been used in studies which seek to estimate the types of taxes which have the most detrimental impact on economic activity as measured by variables such as employment and GDP.

While the HERMES, COSMO and SWITCH models provide the platforms for the ESRI’s most direct inputs into the budgetary process, many other strands of our work are relevant. Some of this work is undertaken as part of joint research programmes with Departments. I will use two examples to illustrate this. Since 2015 the ESRI has operated a joint research programme with the Department of Finance. Under this programme, we have conducted work on the relationship between corporation tax rates and foreign direct investment. The findings were published last week. The relevance to policy debates on taxation should be clear. We are about to begin a project on wealth taxes, using newly available data for wealth which were collected by the CSO on behalf of the Central Bank. I should note that one of the objectives behind this research programme is the generation of greater analytical capacity in the economics of taxation in the ESRI and the Department of Finance.

We also have a joint research programme with the Department of Health. The best known output from this programme was the report on universal health insurance published last year. Under this research programme, we are working to develop a projection model for health expenditure. It is hoped this will allow the Department to assess more accurately the impact of issues such as population ageing and unmet need. The ESRI is also undertaking work on GP usage which will be important in estimating the likely cost of extending free GP care.

The institute is actively engaged in research across a wide range of other areas which are relevant but I will mention just one more for now. For a long time, we have conducted evaluations of training and employment programmes and we continue to do this. The most recent example was an evaluation of the back-to-education allowance and the results were striking, with little evidence of improved employment prospects for participants. The policy conclusion was not that the programme should be ended. Instead, the evaluation prompted questions about how the programme might be improved to meet its targets.

I mentioned at the outset that, from my perspective, one potential benefit from greater Oireachtas scrutiny of the budgetary process will be a greater demand for evidence in support of policy proposals. Putting this another way, I see Oireachtas scrutiny as being another dimension in the move towards more evidence-based policy-making. For this approach to yield benefits, it is important that the evidence used is generated through rigorous analysis. In that context, the ESRI is eager to ensure that the research that we are conducting is readily available to the Oireachtas and communicated directly to members. Our funding model makes it difficult for us to undertake substantial pieces of work unless the work is funded. However, we are eager to co-operate with members and the proposed budget office as part of a general national effort to increase analytical capacity in our public administration with the overall goal of achieving better outcomes for our citizens.

I am happy to take questions.

I thank Professor Barrett for his statement and for his time. He has laid out some of the technical capacity and the two models that the ESRI develops and maintains. Over the past five years I have contacted the ESRI many times seeking budgetary analysis and information on things like distributive analysis. I have used such analysis, as have many other Members, to look at the regressive or progressive nature of recent budgets. What can the ESRI do that others, specifically the Departments of Finance and the Department of Public Expenditure and Reform, cannot? What quantitative and qualitative capacity does the ESRI bring to the table that the Irish Human Rights and Equality Commission will not develop or that the Departments of Public Expenditure and Reform and Finance do not possess? What value does the ESRI bring to the table in the context of this new world of budgetary scrutiny? I hope there will be better budgetary scrutiny. What role does Professor Barrett believe the ESRI should play in this new budgetary regime?

Professor Alan Barrett

In terms of what the ESRI can do that others cannot, my simple answer relates to the level of data analytical capacity within the institute. Let me give some broad figures. The institute has about 100 employees, 60 of whom are high-level researchers. Some 50 of these researchers have PhDs - they are economists, sociologists and psychologists. The ESRI's data analytical capacity is above and beyond that of any other organisation. Part of what we do in terms of using data analytical capacity is generate the sort of models that I have described. For example, our macroeconomic model was developed in conjunction with the Central Bank and a Department of Finance official is working on it with us. The model is, nevertheless, very much led from within the ESRI, partly because we have a tradition of working in the area for about 35 years.

I have outlined data analytics and the generation of models but I shall go a littler further. One of the things the institute is always eager to do is to ensure that a portion of its work is published in international peer reviewed journals. Very often, when one says that to a public policy group, it sounds as if it is a group of academics who are indulging themselves but it is not as simple as that. It is critical, from an ESRI perspective, that we publish our findings in high-quality journals and outlets in order to demonstrate that the work we are doing is of the highest international standards. Some of the organisations mentioned would not have a culture of publishing at a high level. I am not saying that every piece of the ESRI's work must be published in a peer review or anything of that nature. It is very much part of the culture of the institute to be doing the work at a very high level and, of course, from an independent perspective.

A number of organisations around town are research institutes and think tanks. The institute has no ideological backing, background or anything like that. We try to go at the work in as pure a data analytical way as possible, bringing as little ideological baggage as possible. Independence is critical. In summary, there is a technical capacity, a constant drive for quality, an independent dimension and a critical mass argument.

In terms of what the role of the ESRI should be, as far as I am concerned it is an extension of what we do. The institute's mission is to produce research for public policy and to inform civil society, which we take as informing debate. As far as we are concerned we want to continue doing what we are doing. Our joint goals are research excellence and policy impact. We often distinguish ourselves from universities by saying that universities are sometimes interested in research as an end in itself but for the institute, research must have a policy dimension.

To the extent that we are producing work that is of relevance, if it was not relevant we would be very worried. I estimate that between 70% to 80% of the ESRI's work should have meaning for the process of budget scrutiny.

The ESRI's work is very good and is driven by the institute. Therefore, sometimes the work is useful to legislators but sometimes it is not or sometimes we would like extra pieces. For the new committee that will be in place, there will inevitably be a set of prescribed analysis that it will need. Let me give an example. Yesterday, this committee met the Irish Human Rights and Equality Commission. The delegation suggested that equality proofing, be it gender, poverty or human rights proofing, should be done by Departments for various reasons. I suggested to the commission that somebody should play the role the Irish Fiscal Advisory Council, IFAC, plays with the Department of Finance and DPER. In other words, validate the analysis. The IHREC is one group that could potentially build up the capacity to do so. The ESRI is the other obvious group that has the technical capacity to take the analysis from the Departments, put it through the hopper, come back to the standing committee and say, for example, "70% of a matter is great but we question 20% and 10% is flat out wrong." Obviously the ESRI carries out the main distributional analysis that we all use. That role could be easily extended.

Professor Barrett has described the ESRI's ad hoc relationship with the budgetary process. He has said that the ESRI does good work and that the nature of the ESRI's remit means the analysis will, hopefully, be useful to the committee and the Oireachtas. My sense is that the committee will need more formalised arrangements with some analytical body as well as with the Departments.

Does Professor Barrett think that the ESRI is well placed to enter into a service level agreement with the standing committee to say, "Yes, we are going to provide you in January with this, in March with this and in September with the following." Does he believe that the ESRI has a useful role to play in this process? Does he believe that the ESRI can play a useful role in equality, poverty and human rights proofing? Did Garret FitzGerald set up the ESRI to provide a second view apart from the Departments? I heard it was Garret FitzGerald at an interview. Be that true or not, does the professor believe that the ESRI is well positioned to play a quality assurance role for the standing committee when it comes to equality proofing what comes out of the Departments?

Professor Alan Barrett

The Deputy mentioned a range of things and I ask him to give me a second or two to talk about the ESRI's history. The ESRI was not set up by Garret FitzGerald but by Dr. T.K. Whitaker. When I became director of the ESRI one of the fun things that happened was that I got access to a whole load of files on the history of the institute. I could have spent hours going through the files but I disciplined myself not to do so. Garret FitzGerald was certainly involved in correspondence around the time. Let me outline Dr. Whitaker's thinking at the time. The institute was set up around 1960 and formed part of the Lemass-Whitaker revolution.

There was a recognition on Dr. Whitaker's part of a need for greater analytical underpinning to economic policy. The Central Statistics Office, CSO, was generating a great deal of data, but there was not a natural pool of people to conduct the analysis that fed into policy. Dr. Whitaker set it up. He was the leading public servant in Ireland at the time and possibly the greatest of the 20th century. The first director of the institute was Professor Roy Geary, Ireland's greatest statistician of the 20th century and possibly ever. The twin pillars of the institute were research and policy excellence. I am sorry for the diversion, but the Deputy triggered it in my head with reference to Garret FitzGerald.

The Deputy began by saying that our research agenda was driven by ourselves. That is not entirely true. I will take a second to describe how the institute works. We have an annual budget of approximately €9 million, some €2.5 million of which comes in the form of the block grant from the Department of Public Expenditure and Reform. We use this grant-in-aid to perform certain actions, such as the simulating welfare and income tax changes, SWITCH, model, the macro model and short-term forecasting. The balance of our funding - approximately 60% - is partly raised through the research programmes that I mentioned as well as commissioned work. In our research programme with the Department of Finance, the Department makes available a budget of €200,000 or €250,000 per year and a steering committee comprising ESRI and departmental officials agrees the research projects that will be conducted over a two-year or three-year period. It is a rolling cycle of research projects. The Department tells us that it needs work to be done on X, Y and Z and we, as researchers, tell it that data are available to say something interesting on same, there are no interesting data or it is not a tractable problem. There is an iteration. This is how we respond to requests for research. Likewise, we have research programmes with the Departments of Health and Jobs, Enterprise and Innovation and we work with the Health and Safety Authority, HSA, and the Environmental Protection Agency, EPA. Defining the research agenda is interactive.

The relevance to this committee is that there would be the capacity for the institute to engage and determine whether we could provide more tailored research. I hate to mention the grubby issue of money, but the institute's funding model is such that I do not have a team of people at Sir John Rogerson's Quay who have spare time. They are typically doing work under the grant-in-aid or in response to the research requests to which I referred. There is scope to engage with the committee, but a funding stream would need to be attached to it.

Previously, the SWITCH model was used for proofing. Approximately one year ago, a paper considered the gender impact of the recessionary budgets, austerity budgets or whatever one wants to call them. This is a track record on which we could develop.

Typically, we have not been a watchdog for the work of others. I will be very honest and say that we would be slightly concerned about doing that. This might sound precious, but we have a sense of how this sort of work is to be done properly. Consider my comments on the importance of basing work on large-scale representative samples so as to give a real sense of the impact of certain measures across the population. There are ways of doing this. What the institute does is in line with what bodies like the OECD do. I am worried that the committee might set us the task of overseeing someone else's work where we might have a different perspective of how the work should be done. The idea of overseeing someone's work or commenting on it might exist in principle, but I will raise that concern.

It is nice to see Professor Barrett. My understanding is that the SWITCH model is quite old. Does the ESRI have plans to update it? I appreciate that this would require funding, probably from the Government. Some work has been done on a more real-time model.

What would Professor Barrett describe as the strengths and weaknesses of this model? For instance, certain distributional factors are included while others are not. If a deprived area gets a new school, it does not count under the SWITCH model even though the impact on the children's lives is significant. There are other, more concrete, examples. I have never fully understood the rules of what the ESRI counts in and out. Will Professor Barrett give us a picture of the situation? His reference to the back to education allowance was probably a good example of the difference between researchers and politicians. A great deal of money is spent on the allowance annually, amounting to more than €200 million in the Department of Social Protection. The ESRI's report was critical of the allowance because it did not seem to lead to people immediately getting jobs. The response from many Deputies was that the allowance was important because it could, for example, be accessed by lone parents and the educational attainments had a value in themselves for a lone parent and his or her children, given that they are the people most at risk of poverty, but this is not captured in some of the types of analysis conducted by the ESRI. This follows on from the previous question. While I appreciate Professor Barrett's point that the ESRI does not have the job to be political, even with a small "p", it is analytical.

Down the years, I have seen a particular criticism of something that has led to a divide between the political perception and the analysis, namely, community employment. As a consequence, when the crisis hit Ireland, everyone with a briefcase who came to advise us from outside read all of the ESRI's reports, which were not flattering towards the likes of community-based employment schemes even though many Deputies from across the political spectrum favoured them strongly because they saw the local impact. Being able to put forward a range of views is important, but the ESRI could play a valuable role for the committee in explaining matters from an economic perspective. The institute's staff are mostly economists and social researchers whereas we, as a political committee composed of diverse elected politicians, might view matters differently.

With the exception of the Department of Social Protection, I am unaware of any Department conducting pre-budget and post-budget distributional analyses of the budget on a continual basis, yet it is one of the suggestions that has been discussed at this committee. Does Professor Barrett believe it is feasible for a range of Departments to conduct these analyses? The Department of Social Protection's social impact analysis is partly based on the ESRI's SWITCH model work. If it could find its way into some of the committee's analysis, it would be useful, but we would probably need a great deal of technical advice as to what would be desirable to be received from Departments so that, in theory, a budget's distributional impacts generally and on specific Departments could be analysed broadly. On budget day, the initial pre-budget impact analysis could also be published.

This was mentioned by a lot of people in the discussions we have had to date. Does Professor Barrett see that as being feasible? Would the ESRI have a role to play in assisting the Departments and the budgetary oversight committee in that analysis?

Professor Alan Barrett

The Deputy has made a number of points which I will run through. On the point about the SWITCH model being outdated, the SWITCH model is continuously updated. The dataset used is the European Union survey on living and income conditions, which is collected in Ireland and throughout the rest of Europe and that data is constantly fed in. When I made the point that the model was developed and maintained, a significant amount of the ESRI investment on this, using the Department of Public Expenditure and Reform grant-in-aid, is in ensuring that the data is up to date. We have had discussions with a range of Departments to see if the funding for SWITCH can be increased to allow for a quicker updating and also for the pooling of data across waves. We are in a pretty good ---

On that point, what would Professor Barrett prune and what additionality would the ESRI provide if it got that funding?

Professor Alan Barrett

The up-to-date part is simple; the pooling across waves less so. The survey is collected every year. If one starts adding waves together, so that one is not just dealing with one year's data in the model but two or three years' data, that means that one has much bigger sample sizes and can start looking at, for example, the effect of welfare payments that are received by a relatively small group of people. I can give one example that came up recently, which I hope is a good example. An examination of the effect of housing-related payments was difficult with the original SWITCH because such payments are not like unemployment or pension payments. The number of people involved is smaller and therefore the data was not so robust but the additional data work that has happened there is going to allow for richer analysis.

Deputy Burton asked a broader question which is an interesting one, on the advantages and disadvantages of SWITCH. I have already touched on the advantages in terms of the fact that there is simply no other way to get that population-level sense of the distributional impacts if one is not doing SWITCH-type analysis. That advantage is clear. The downside, of course, is what is not included although we have tried to extend things. One of the big innovations we are working on at the moment is the issue of medical cards, for which I should give my colleague Dr. Tim Callan and his team credit. The issue of medical cards has now come on the agenda. The Department of Health will become a co-funder of the SWITCH model with a view to factoring in medical card entitlement and impact. That is a very significant addition -----

That makes a huge difference and people know that. If the person living in house No. 1 on a street has a medical card while the person living in house No. 10 is working in a low-paid job but does not qualify for a medical card, on the SWITCH model, the only income for the person in house No. 1 that is included is that from the Department of Social Protection but the value of the medical card to that person, who may have a spouse and two children, could be €400 to €500 per year, particularly if there are medical conditions involved.

Professor Alan Barrett

To be honest this is a good example of positive interactions across a number of Departments. For a very long time the primary funder of the SWITCH model was the Department of Social Protection which did not go near things like medical cards and so forth. Now that more Departments are getting involved, those sorts of issues are being included. The other thing that SWITCH was previously not very good at was capturing issues around indirect taxes. This is very important because we know from research that I and others have done that indirect taxes are typically more regressive than direct taxes. Therefore, when we were doing the analyses but not picking up the indirect elements, there was an omission. That said, these are weaknesses that can be plugged and there is a very active programme of work under way to make sure that they are plugged.

There is a bigger criticism that can be made. The critical thing that SWITCH tries to get a handle on is distributional impact and whether budgets are regressive or progressive. One could make a budget look very progressive by introducing, say, a 98% tax rate at the higher end and increasing all welfare payments by 100%. That would be a very progressive budget ---

That sounds good to me.

Professor Alan Barrett

Yes, I can imagine. I just glanced in that direction and saw a smiling face.

That would look good from a distribution perspective but some people would question whether a budget like that would have economic impacts that are worth discussing. At that level, one wants to have interactions and parallel analysis using the macroeconomic model and the SWITCH model. I have heard people criticise SWITCH on that basis, namely that it can give one a particular outcome but that it is partial.

The Deputy mentioned our analysis of the back to education allowance scheme and criticisms of that. Just to be clear on that, the Department of Social Protection asked us to do a study to determine the employment effects of the back to education allowance. The ESRI was asked a very clear question and had a contract to deliver on that. We answered that question, which was that the allowance had very little impact on people's employment. We said in the report that there were multiple reasons for having such a programme and that it may be doing spectacularly well in other dimensions but we were not asked to look at those other dimensions. If one thinks in terms of a committee, I can imagine a politician who is very enthusiastic about the back to education programme and believes it is wonderful because it does A, B and C, where A might be generating good pathways back to employment. I believe there is value in even partially looking at and interrogating whether programmes are doing what they claim to do. There was a similar issue with regard to community employment schemes. For a long time there was a narrative around community employment that suggested that it was an employment programme. I could say that the ESRI said but the ESRI never says anything - ESRI researchers write their reports and are responsible for the content. The ESRI simply publishes material and gives it a stamp to verify that it has passed refereeing processes and that the institute believes the analysis is sufficiently robust. The ESRI itself never says anything. On community employment, the researchers said that the scheme was not an effective way of getting people back into mainstream employment. That was all.

The last point I would make is that one of the research programmes currently under way at the ESRI is with Pobal. To come back to an earlier theme about research and why it should be more interactive, the ESRI has being doing evaluations with Pobal. That organisation has raised a very important point with the ESRI. We have been looking at the effect of Pobal programmes on individuals. Let us say for example the question is whether a particular Pobal programme helps to get people back into the labour market. However, Pobal is very interested in the community-level impact of what it is doing. Pobal has come back to the ESRI, told us that what we are doing is really good but have asked us to develop tools and mechanisms that will allow for an assessment of whether certain programmes are having a community-level effect. We are now talking to Pobal about how this might be done. That is a really positive example of us producing research output, engaging with an agency which asks us to do more and then the ESRI trying to do that.

The last issue raised concerned various Departments doing lots of analyses-----

That could be in relation to the capital programme or any major amounts of public spending.

Professor Alan Barrett

My only concern about a multiplicity of analyses is that if people are using a whole range of different models and assumptions it can be chaotic. A budgetary committee would need information that is useful and not just bits of information that are supportive of particular positions. There is an argument to be made for ensuring that any budgetary oversight committee which will be discussing the evidence on a particular policy proposal will at least agree on the single piece of evidence that it wants to deal with, as opposed to having multiple pieces of evidence. That could be tricky however.

We have three Deputies offering and several other witnesses to deal with so I urge members to ask questions rather than make statements.

I have a number of short questions for Professor Barrett. Does the Department of Finance rely on the ESRI's model or does it have its own model? What model does the European Commission use? How does the ESRI relate to the European Commission in terms of any modelling analysis the latter carries out? Is the model we use - set up by T.K. Whitaker, which was an interesting piece of history - matched by other similar sized countries in the European Union, like Austria, Finland or Denmark?

What do they tend to use in terms of economic research analysis? How much of the ESRI's research analysis, particularly that relating to the Department of Finance, takes place ex ante rather than looking at decisions after the fact? As a specific example, what involvement did the ESRI have in analysing the spring economic statement, which was published yesterday? What involvement did it have in analysing the earlier stability programme update before it was published? I was interested to hear that the ESRI has 100 staff. How many staff did the ESRI have five years ago and ten years ago?

Professor Alan Barrett

I will try to give focused answers to the Deputy's focused questions. The Department of Finance does not have its own macroeconomic model of the Irish economy. It has historically relied on the ESRI model. To be clear, it made an input into the development of the original HERMES model approximately 30 years ago. An official from the Department is working with us on the development of the COSMO model. It does not have its own model. We lead the model that is used, although it has an input into that model. I am aware that part of the Department's rationale for having an official working closely with us is that this person can do their own runs of the model. Typically, we have done those runs for the Department. On the question of how countries internationally organise their------

I also asked about the European Commission model.

Professor Alan Barrett

I will be honest and say I am slightly out of date with regard to the European Commission model. We do not tend to engage in a model-on-model sort of interaction with the European Commission. When Commission officials come to Dublin to do assessment exercises and forecast rounds, etc., they meet representatives of the ESRI and we discuss the output of our model. The Commission certainly has its own models. I think one of them is called the QUEST model or something like that. I am not necessarily aware of conversations between the ESRI and the European Commission on modelling issues specifically.

My understanding from previous years when I was more directly involved in these issues was that the Commission was very critical of our modelling. My understanding was that the Commission had very different views about how modelling should be done.

Professor Alan Barrett

I would not quite share that perspective. Funnily enough, we were very critical of some of the things done by the Commission. Needless to say, there was a general recognition during the economic crisis that everybody's economic models had performed very badly. One group might have been saying its model did not perform quite as badly as another group's model. I might be missing something, but I am honestly not aware of ongoing debates on that issue.

Professor Barrett has suggested that there is no difference in this area, but my understanding is that the stability programme update revealed that we have a difficulty with the different modelling approach of the Commission in terms of some of the economic analysis of where we are.

Professor Alan Barrett

The Deputy might be referring to the very particular issue of the methods used to measure the concept of potential output. This was a sort of academic concern for a very long time, but it has taken on a new reality under the new fiscal rules. While the Deputy is right when he suggests there was criticism regarding a specific issue, I do not think the criticism necessarily related to broad modelling issues.

Deputy Ryan's second question related to economic research in other countries. There is no single model on this. Some of the larger countries have multiple models with various different sources of funding. The United Kingdom has two research institutes that mirror the ESRI. Both of these agencies - the National Institute of Economic and Social Research and the Institute for Fiscal Studies - tend to draw on Government funding through competitive grants. The UK system is a little different from the system in Ireland. The CPB in the Netherlands is an independent but fully state-funded statutory organisation that does everything the ESRI does, as well as many things that are proposed for the budgetary office. For example, it does costings of budgetary programmes and election manifestos, economic forecasting and a range of other activities. There is no sort of single model that applies.

The Deputy also asked whether our analysis is ex ante or ex post. The vast majority of what the ESRI does would have an ex post dimension. When a particular policy or programme is operating, we typically use data to see what is, or what has been in the recent past. The exception to that is our input into something like the spring or summer economic statement. I will explain how it works.

The Department of Finance asks the institute to use the macro model to run a whole load of simulations. I touched on this in my opening statement. A simulation might relate to the impact on Ireland of GDP in the global economy being 1% lower than it is or the effect on this country of an interest rate increase. The Department will have a sort of central forecast on which all its numbers hang and we look at the possible effects of various sensitive-type issues on that. The Department asks us to do policy simulations and to ask "What if" questions. This requires us to look at the effects of major tax changes, for example. Capital expenditure often arises in this context. We might be asked to assess the impact of a doubling in the public capital programme, for example. I will explain how it tends to work. After the Department of Finance has contacted us, one of my colleagues will work quite intensively for a week or two. It is very often the case that just one or two of the simulations run by the ESRI make it into the budgetary documentation. There is a sense that a lot of thought goes into this work, but the set of material that appears is quite condensed. There would be a good degree of ex ante work at that level.

Historically, the European Commission was one of the biggest generators of ex ante analysis. In the era when a great deal of EU money was being spent in Ireland, the EU was very good at demanding that we did ex ante evaluations of moneys that were to be spent. The ESRI worked on these programmes at a sectoral level but also at a macro level. I started at the ESRI in 1994. One of the first studies I worked on was published in 1997. The project was led by Patrick Honohan. We put together figures as part of one of the major evaluations of all EU structural spending in Ireland in the 1990s. There was great discipline in those exercises because they forced Departments and others to think through what the effects would be and whether the sort of expenditure that was being talked about could be justified. I have often thought that one of the unfortunate things about European money drying up, apart from the money going, was that some of this intensive evaluation culture left us. It is still there, but it is not applied as rigorously as it previously was. Others have spoken about this as well.

The Deputy also asked about staff numbers. While I can provide the numbers, this issue is slightly complicated by a particular decision that was taken by Senator Reilly when he was Minister for Health. The ESRI had a significant number of people - between 20 and 25 officials - collecting hospital-related data. The then Minister decided that work would be taken from the ESRI and given to what ultimately became the Healthcare Pricing Office. We lost 25 people, in a sense, at that time. We have been building up ever since. I could give the Deputy figures, but they would not illustrate neatly the underlying situation in the ESRI. I should say the ESRI was not subject to the public sector employment embargo that affected other offices. Its independence meant it could carry on hiring in recent years. Its grants and pay levels were cut and various things happened, but it was not subject to the employment embargo.

Cuirim fáilte roimh na finnéithe. I will start with a couple of quick questions. The HERMES model did not include the financial sector. Has that been rectified?

Professor Alan Barrett


Okay. Regardless of what kind of budgetary committee is set up, it will have to deal with the spring economic statement in some way. I will not ask for the ESRI's view on this because it does not have a view. What is Professor Barrett's view on how we deal with that data? Sometimes there are conflicts between what the Irish Fiscal Advisory Council tells us, what the Department of Finance tells us and what the Government then presents in the spring economic statement. It was announced yesterday that there will be approximately €660 million available for current expenditure. That does not, however, take account of the standstill position, on which the ESRI has commented in the past. I think it said, in the context of last year's budget, that welfare increases of €450 million would be needed to manage the standstill position, which is based on indexed payments. It seems there is some kind of consensus on moving beyond indexation of tax bands or social welfare payments. I think this is a way to suggest that there is more flexibility than is actually the case.

In terms of proper budgeting and dealing with the evidence and data, is that an appropriate stance to take or should the standstill position or indexing be factored into the way data are presented to us?

Professor Alan Barrett

I dealt with the HERMES issue earlier. The development of the new model partly related to the need to plug that particular gap. The financial side, therefore, has now been factored into the model.

I am not sure there is a right or a wrong answer to this. The institute has typically in its analysis indexed all welfare payments to wage increases in its thinking about the budget. That was the standard approach we took. It may be partly driven by the fact that the institute for a long time was involved in the measurement of poverty and reporting on poverty statistics. It is a simple point that if wages are increasing in the economy and welfare payments remain constant, increased poverty is being engineered into the system given poverty is defined as a proportion of average wages. Researchers in the institute always had trouble getting their heads around a scenario where there was non-indexation while wages were increasing, but they would have to stand over a position that poverty could be increasing when it was a clear objective of Government to reduce it. It always struck me as somewhat inconsistent to say there should be zero indexation but any indexation is a proactive decision to provide. There are different positions on it. I have had this discussion with colleagues in the Department of Finance who always make the more legalistic point that when the Minister stands up on the day of the budget, it is a decision to increase and it is not automatic. There is not necessarily a right and wrong answer but the key point is the committee needs to have analytical capacity, either in the heads of members or provided to them by someone in a support role in order to put them in a position to tease out the different positions that will be presented.

These are complicated issues. The worst of all worlds is where people are not explicit about the assumptions underpinning the figures they are talking about. To talk delicately about the issue that blew up during the election campaign in respect of fiscal space and precisely how much money is available, nobody was necessarily saying anything that was incorrect but people were not being completely explicit about the assumptions underpinning the figures they provided. There is a sense that members need to be well informed enough to know what are the questions they should ask.

I agree with Professor Barrett and I also agree with the Department of Finance officials that it is a Government decision to index tax credits or, indeed, social welfare payments. Presentation of data is crucial. Professor John FitzGerald's comments to the banking inquiry were insightful. He said models do not get the answers wrong; it is the individuals who use the models or who do not understand them that get the answers wrong. Giving a model to someone who does not understand it is similar to giving a box of matches to a child. There is an expectation in society that approximately €600 million will be available for current expenditure. Should it not be the case that - while it is ultimately a Government decision - a standstill position would require X amount of this to be absorbed?

We are well familiar with what the ESRI does and we have our ups and downs in respect of some of its analysis. It is interesting that the question that is put sometimes reflects the way the answer is given. Professor Barrett has experience of the Irish Fiscal Advisory Council, the ESRI and the national economic dialogue. If he was to write the report we, as a committee, are tasked with, which is to make recommendations on how a budgetary committee should be established and the establishment of an independent parliamentary costings office, what priority recommendations would he make?

Professor Alan Barrett

That is a big question. To be succinct, the quality of the information provided is critical.

Is it Professor Barrett's view that in advising the committee the Government hopes to set up, the demographic pressures and the impact of inflation on the fiscal space and so on have to be spelt out clearly as a matter of course in the discussions around the budget?

Professor Alan Barrett

It is just one of the things that needs to be spelt out.

But it should be spelt out.

Professor Alan Barrett

The assumptions underpinning whatever is done should be spelt out and should be clear but it is only one of the issues. I do not know what vision members have of how this will pan out. The potential added value relates to the fact that the budgetary process is carried on behind closed doors to a certain extent at the moment. Members will be in a position to publicly ask people to justify whatever choices they are making. To interrogate, it comes back to information and the quality of the committee's investigation will be heavily dependent on the quality of the information that members have to hand.

I have talked about what the ESRI does. We do a good job at providing good information. It may be partial. For example, it may be that we have been asked to study something in a particular way but it is based on good, rigorous analysis. Members need good quality information that they can use but they should not become totally fixated that the only issues floating around are demographic pressures and inflation. The budget relates to between €50 billion and €60 billion worth of expenditure and taxes and so on. At the National Economic Dialogue, people constantly said we cannot raise tax on labour because that is bad for jobs. Perhaps that is true. For example, if 1% was added to the marginal tax rate, what would that do to jobs? That is an empirical question. Sometimes, a mantra is put out there. Somebody like the good Deputy might have suggested that corporation tax should increase and perhaps it should, but at least the ESRI published analysis last week which showed that if that was done, this is what we estimate the impact would be on foreign direct investment. The Deputy may say that is a price worth paying or he may disagree with the analysis but at least he has a piece of work that tries to set out and answer the question and provide evidence on the issue he is dealing with. My vision of what the committee should be doing is bringing public forum evidence to bear. When the evidence is not there, members will ask people to justify the choices they are making.

I agree completely with that. I welcome the entire process and it is good that it is being opened up. We have to figure out how best to establish that process to achieve the maximum transparency and the most comprehensive debate in the run-in to producing a budget.

Professor Barrett made a point about not having ideological baggage and so on. One of the issues Nyberg pointed out in his report was that contrarian thinking was not taken seriously. That was an understatement in the context of the failure of all sorts of models and think tanks to see what was coming in 2008 and to understand how severe it could be. ESRI staff put their hands up and said they did not recognise the extent of the banking and financial crisis. How can we be sure the same is not happening again and that contrarian thinking is still not being marginalised within the ESRI or anywhere else?

When I studied economics for the leaving certificate, the entire book by Noel T. Palmer was about market theory. There was one page on Marx's labour theory of value and it just said it was wrong.

They just went on to enunciate the Smith and Ricardo outlook. Can Professor Barrett assure me that the contrarians get a fair shot?

Professor Alan Barrett

I will take two stabs at explaining this matter.

It is handy to have a professor of economics present.

Professor Alan Barrett

Let me take my first stab at explaining. I concede that the institute did not see the full extent of what was coming down the road but we published Professor Morgan Kelly's paper on housing. I know that it was Morgan Kelly's and UCD's work but the institute published the paper. Why? I remember that he submitted the paper to our quarterly economic commentary - we sometimes publish work by outsiders. We read his paper and thought it was an interesting and important piece of work. We did not agree with the findings but we published it because we thought it met the type of analytical standards that the institute appreciated. In that context, surely we deserve some credit for having done so. We came under an awful lot of pressure about publishing the paper but it was important, from the independence perspective of the ESRI, that it was published.

Who put the institute under pressure?

I was going to ask the same question.

Professor Alan Barrett

The Department of Finance objected, at the time, to us publishing it. I can develop the point even though I have given a simple answer.

I will take a second stab at explaining. As much as 98% of what we do is analysing data sets. We do not sit around all day theorising about market economies, Marx and all those sorts of things. Typically, we get a data set and then we figure out what the data is telling us about whatever issue. Let me give the example of a policy evaluation of the back to education allowance. In that case one has data and one tries to hammer it out.

Everything that comes out of the institute is peer reviewed. That means two or three people within our building will read people's work. We often send it to outside reviewers as well and say, "Is the analysis of the data up to a good standard and are people drawing the correct conclusions on the basis of the results, as presented?" I cannot give the committee a cast iron 100% guarantee that everything is always kosher because people are human. People can make mistakes in data analytics and people's interpretations can go wrong. What I can say is that the institute has tight quality control devices to make sure that the analysis of the data and the presentation is as truthful as can possibly be imagined, which to me is pure. We are not funded by any particular people on the left or the right. We are, on occasion, criticised with people saying, "You are funded by the State so that must mean the institute is at its beck and call". To explain, I refer to the back to education allowance or the publication of Morgan Kelly's paper. I always have the view that the institute does not set out to annoy anybody in Government circles. I always think that there is something healthy about being criticised because if we did not fall out with Departments periodically, and if they did not like what we said, then we would not be doing our job. The Department of Social Protection has accepted what we have done. Various agencies over time, if one produces results that are very critical of them, tend not to be too pleased, needless to say, but that is the institute doing its job.

I appreciate that but I want to plumb into this matter a little further. The professor has said that the institute deals with data and has certain models, which implies that there is not much in terms of subjective analysis. Is that what he is saying? Is there debate?

Professor Alan Barrett

There is intense debate. People are really put through their paces within the institute. No piece of work comes out of the institute unless it has been reviewed by two or three people. We are the Economic and Social Research Institute so we have a pool of sociologists who perhaps have different disciplinary perspective on matters. I am not saying we get it perfect or anything like that. There are notions that sometimes exist about us and the Deputy may have heard better descriptions than me. We are not a right-wing organisation.

I did not assume that. I want to understand the process of producing analyses and reports. Is the full spectrum of economic thought fed into the analyses? Is everything just a compromise between different people?

Professor Alan Barrett

The astute thing for me to do would be to send the Deputy our boring old publications policy that would give him a sense of how we do things but I want to explain it in a much better way. I had the following conversation with somebody from the Department of Social Protection who once attended one of our review meetings; I will not name the person. She said she was absolutely amazed at how robust the discussion was over a piece of work. I refer to a situation where a researcher has produced a piece of work that has really been kicked around and asked to justify the conclusions reached. The Department had funded the work and one of its officials attended the refereeing meeting. She commented on how robust the debate was. She then assumed that half of the group would not talk to the other as they left the meeting but realised that everybody went to coffee together. Robust debate is considered part of one's role in the institute. I often say that the people who work in the institute have a role to play in terms of writing the reports but they also have the important role of commenting on other people's reports. I shall outline one of the areas where this arises, and I may meander ever so slightly. People say that the ESRI could be run virtually, that we do not need to have an office and everybody could work in their own office but send their work in. However, another important dimension of the ESRI is its ability to have a collection of people who work together, interrogate people's work and feed off one another.

I am nearly out of time so will ask a few bullet point questions. I am glad, and I like to think it is due to the presence of the new left joining the Irish Parliament, that the institute is working on wealth taxes. How close are we to having an analysis of the possible impact of different models of wealth tax?

I shall paraphrase the Deputy, this is the arrangements committee for the ultimate budget committee. His question is not related to arrangements but is more suited for the committee when it is established. I want to let Deputy Broughan contribute and there are two other sets of witnesses.

Professor Barrett introduced the issue in his statement.

He did and we will get a chance to talk to him again about the matter once the committee is established. Deputy Barrett's query is not an arrangement matter.

It is in the sense that there are certain things like-----

I have a copy of that book here some place that proves the Deputy studied and presented economics.

-----poverty analysis and all that kind of stuff, which were not in previous budgets. A big lacuna in the-----

The Deputy wants to know if the ESRI has completed the research.

How close is the ESRI to completing the research?

I urge Professor Barrett to pick one sentence.

I ask Professor Barrett to bank my question as I have another question. I am afraid that the models do not take enough account of human experience, which kind of follows Deputy Burton's comment. Should the models include surveys of target groups? The people affected should be asked how something will impact on them, rather than just use statistics and figures. I often think statistics and figures can be limited in what they say about the impact on human beings, which is after all what this is really all about.

Professor Alan Barrett

In terms of when the wealth tax work will be available, the Department of Finance holds an annual taxation conference in November. I cannot remember the precise date but I know a presentation is scheduled. I will make sure an invitation is sent directly to the Deputy.

The Deputy does not understand that even economists have feelings but I will not rehearse the discussion we had on this once before. I agree with him that only so much can be captured in a model. He needs to understand that economists are, increasingly, looking at a whole range of indicators, including notions of well-being. It is becoming a big part of what we do and I shall give a personal example. My most recent economics paper is on the mental effects of retirement and was published a few weeks ago. In terms of what was the dependent variable that I was interested in, and we have used Trinity's TILDA data, The Irish Longitudinal Study on Ageing, we are looking at whether people's depressive symptoms increase when they retire.

I read in a press release this morning that there will be a paper next week on issues and indicators of quality of life. The institute is doing a lot of work and not silent on these issues. Including quality of life issues in a macro model is a big project and I hope we can ultimately do something in that regard. We are active in the areas in question.

I welcome Professor Barrett and his colleagues. Does he envisage the ESRI not having a formal relationship with the new committee as such? Does he see it giving information or advice at particular intersections in the fiscal and budgetary cycles without, in fact, having a formal relationship? Would the institute need significant additional capacity such as 100 staff to adequately assist the committee and resource it if a formal relationship was required? It is now significantly bigger than it was before the crash. There was criticism of it in the run-in to the crash.

Professor Barrett has stressed that data are crucial. I had a similar experience to that of Deputy Richard Boyd Barrett when I brought before the last Dáil my High Pay and Wealth Commission Bill which the Government refused to advance. At the time, I was of the view that there were large gaps in the data available, especially data indicating the level of interaction with Revenue. Is there a web of information which we do not have but which other countries’ budget committees do? In Norway everybody knows what everyone earns and how much wealth everyone has. Is that something we should try to advance?

Professor Alan Barrett

A formal relationship between the ESRI and the committee would be possible and is something we can explore. There would be no problem if the relationship involved relatively minor inputs such as ESRI staff presenting on various issues. That is part of our public good remit. If the committee wanted to have pieces of work done, that would be possible and we would be happy to engage in those discussions. I hate mentioning money, but that is an issue that would accompany it. I am not allowed to go back to headquarters unless that message is well understood. We can have discussions and there are possibilities.

There is an issue as to whether data are being collected. The situation in Scandinavia is different because there is the tracking of individuals. I attended a presentation recently with the head of the Danish statistics office. He said their census was run by two guys and that it took them about one week to complete. They do not have to go out and count everybody because they have data for everybody. They have phenomenal data which are collected over time. They are not just for their income and wealth but also their engagement with the health system and so forth. They have very different ways of doing it for historical reasons. We do not have that level of data available. However, more data are being collected over time. The wealth issue is one on which we typically did not collect data in Ireland, but that gap is being plugged through the Central Statistics Office and the Central Bank. From an ESRI perspective, we find the CSO very accommodating. Subject to the rules of data protection, it generally tries to facilitate us with data it has available. If the data are available, we have good access to them.

Professor John McHale and his colleagues will talk about their more direct interactions with the Department of Finance.

I thank Professor Alan Barrett for his attendance. I am sure that when the committee is established properly, we will be back in touch.