Mr. Peter van de Ven, the head of the national accounts division at the OECD. As such, he is responsible for the statistics on National Accounts, Purchasing Power Parities, Prices and Composite Leading Indicators
• S: You talked about the special media attention devoted to indicators such as government debt and deficit. There is currently a certain “competition” among OECD countries to improve indicators. Which effects have this media attention on your work as head of division of National Accounts?
Van de VEN: That's not an easy question and there are many ways to look at it. If we have a look back at my 30 years of experience I can see a couple of developments. First, the media attention for our data has increased significantly. You can also see that there is a cry for more and more recent data. When I started in National Accounts (NA), our first results were published six months after the end of the year and were related to annual data. Now it's all about quarterly data, so there is a certain pressure to come up with more and more timely data. Nowadays, within Europe, for example, first estimates for quarterly growth are published six weeks after the end of the quarter (so from six months after the end of the year, to now 1.5 months after the end of the quarter). This has also to do with the growing importance of data in general, NA data in particular. There is a lot of attention and I think in general one can say that the quality of data and its reliability has improved considerably. This is most certainly true for the international comparability of data between (European) countries, among others because of international standards. There are also a lot of controls and procedures. But on the other hand, this pressure to come up with more and more timely data necessitates relying on less resource data. That means that you have to make more estimates. Statistical offices usually review the difference between the first estimate and the final estimate. The final estimate usually is compiled only 3 or 4 years after the relevant period and what you see is typically that the first estimate differs from the final estimate in a range of +/- 0.5 %. So, because of this difference, if people attach a lot on e.g. minus 0.1%, I say: let's see what we get when we have the final estimate. This difference is actually not so bad, let's say in a range of +/- 0.5%, it becomes worse if you have a systematic over- or underestimation. Let's take the Dutch example: what we saw, when we made comparisons for the last 10-15 years, was a systematic underestimation. The first estimate was an underestimation of 0.5% on average, our range was roughly between 0 and 1%. What you do then is you look closely at your estimation methods and you try to improve them in order to get rid of this systematic bias.
• PH: Is this media attention typical for Europe or do you observe the same phenomenon on other continents?
Van de VEN: No, the same is true all over the world. Actually, the U.S.is producing even quicker estimates: like the UK they produce them 30 days after the end of the quarter, but often they can have significant revisions later on. In the U.S., the data on economic growth are far more important, than the ones in Europe because they also directly affect stock markets. So when they compile their estimates, they make sure that nothing gets out before it is published and everybody gets the data at the same time and they physically lock the room and journalists may even come in to be the first ones to see it but they cannot get out.
• S: You mentioned during your talk the Excessive Debt Procedure (EDP) within the EU Stability and Growth Pact that focus on government debt and deficit. You said it would be better to focus on net savings and on net debt instead. Do you think that these alternative indicators would change policy proposals, that they may have different effects on the resolution of the crisis?
Van de VEN: Perhaps not the change to net savings and net debt as such, but looking at a broader picture would perhaps have an impact. Now the focus is almost entirely on this gross debt and deficit, and that seems to have some impact on policy as well. In the presentation, I mentioned the example of Eastern European countries, where they had a pension reform. They tried to set aside part of the pension contributions in a separate fund and to create a real fund for the payment of future benefits, which seems to be a very sensible policy from an economic point of view. However, that also had a negative impact on government deficit and then you get closer to this benchmark of 3 percent. When the crisis started with the accompanying pressure on government deficit, some of them decided to stop this economic policy and even reverse it. The latter actually had a very positive effect on deficit. This change of policy is a pity. If you would take a broader picture and take into account what we call implicit pension liabilities, which are not recognized as debt in the system, if you would take those into account, then you probably wouldn't have this negative impact on policy.
• PH: Do you think that the Euro debt crisis enforced the discussion whether indicators such as debt, deficit as well as GDP growth are appropriate or did this discussion start already before?
Van de VEN: I think that this discussion already started before. The relevant indicators have been chosen at a political level at a certain time after a lot of political debate. But they finally agreed about it. Afterwards, we had something close to a “crisis” with at the front France and Germany crossing the three percent barrier, and a lot of discussions. If you open up this agreement, you would open up a box of Pandora and I think politicians are not inclined to do that. They have agreed on it: so, let's stick to it. However, because the indicators are so heavily used, so strongly politicized, they lose some of their usefulness.
• PH: Why do you use at the OECD another definition for debt as in the Maastricht Criteria?
Van de VEN: It is the other way around. For NA, you have international standards. They are called the System of NA (SNA); the one, which is now applicable, is the SNA 1993. Based on these standards a European equivalent has been produced, the European System of Accounts (ESA) in 1995. The standards are the same as the one in SNA 1993, but it is in some respects more precise in its definitions. Basically that's the system one has to apply for the calculation of NA in the European Union. When they set up that agreement on deficit and debt they did not take for debt, for example, the full definition, as it is included in the core system of NA. They excluded, for example, other amounts payable, they used a quite specific definition, we call it the Maastricht definition. The Maastricht definitions are pretty close to the NA definitions, but they are not fully the same. We have now changed the standards to SNA 2008 and ESA 2010, that will be introduced in Europe in 2014, and they try to bring the definitions together as much as possible, because it creates a lot of confusion. Take, for example, government deficit. There is a difference between the deficit according to the SNA/ESA and the Maastricht deficit related to the recording of derivatives. Everybody gets confused and you want to get rid of it. Hopefully we can achieve that.
• PH: What about the structural deficit?
Van de VEN: You could look at it in different ways. Eurostat for example is looking at having seasonally adjusted quarterly deficit figures. That's one thing; you can imagine that if you look at quarterly deficit, it bumps up and down with taxes collected in a certain quarter, so you want to remove those cyclical movements. But you probably refer to deficit figures which are adjusted for business cycles, output gap movements. I'm not really involved in that line of business.
• PH: In each Euro Area member we have currently a national statistics authority depending on and supervised by the government in power. Do you think that a more independent model could lead to an improvement of the quality of statistics delivered to Eurostat (and the OECD)?
Van de VEN: Actually you have different models in the EU and also outside Europe, and sometimes you have very formal arrangements, which make the statistical office or statistical authority more independent. Consider, for example, the Netherlands: the statistical office is connected to the Ministry of Economic Affairs but it's a separate legal entity. The Ministry of Economic Affairs cannot influence the statistical program. In France you may have a different model, but INSEE is also known for being independent. When you look at Greece, one of the first things that was done after the events in the end of 2009 was to create a legal structure in which the statistical office would become more independent. But, of course, you can create a nice legal structure, but still it depends on how people interact. It is also a part of culture. In the Netherlands for example it is very much accepted that we take an independent view and the Ministries also respect it. Of course they sometimes also try to influence but it is clear that we take the decisions in e.g. the recording of government transactions, and it is not someone else. But you can have a different culture. Let's assume that the Ministry of Finance of the Netherlands wouldn't be happy with what I do as director of national accounts, and they would try in some way to get me out of business, and I subsequently would go to a newspaper, I wonder who has the biggest problem: I or the Ministry? I don't know how it is when something like that would happen in other countries. You must realize that is all part of the game that the Ministry of Finance tries to influence things. On the other hand, as a statistician, you have to be firm and act independently.
• S: If you cross the three percent thresholds, or just below, it may be easier to be strict with definition of independence. Do you think that for example in Greece, a higher deficit is correlated with higher political danger?
Van de VEN: It’s perhaps also part of economic culture. I take the example of Germany where some time ago you did have a debate on how to move forward: we are not as competitive as we used to be, what should be done, how should we address this? You had very moderate wage increases for a longer period of time, which did improve the competitiveness over time. To address these issues in the political arena and to find a solution, that’s very important. I have the feeling, however, that it becomes more and more difficult. This is a more general observation that doesn't have anything to do with statistics as such: we need more substantive political debates on fundamental issues, how to address and solve them. That is very difficult in an environment where much is very focused on the short term and people tend to think how does that affect my income or my situation. These fundamental issues are real societal issues that need to be addressed. It needs a lot of political courage to address issues like the ageing society and how to deal with it. You see that sometimes happening under the pressure of an economic or financial crisis, like the current discussions about retirement age, but it is not easy.
• S: What are in general the job opportunities at the OECD for young economists, and especially in your department?
Van de VEN: Let me first say that a former TSE student is actually working in my division. To enter at the OECD you would normally apply for a job and it would be good to bring some experience in the relevant field, since the competition is quite substantial. People like to come to work at the OECD. Another way to pursue this would be an internship of e.g. 6 months. By this, people see the work you are doing, you can learn on the job and collect experience. If you perform well and there is the need for an auxiliary you may be asked if you would like to do this for some period of time, and then there is perhaps a job opening where you could apply. Your chances for getting a job in the OECD increase significantly if your future employer knows you already and knows that you very capable.
• S: Would you rather say that the OECD is a job opportunity for people recently graduated from university?
Van de VEN: There is Young Professionals Program (YPP), which is an OECD broad program. There are a lot of applications for that, usually only the best people are taken, and the best people are usually those who have a couple of years experience in the field for which we look for young professionals. Sometimes they bring PhD etc., so we are really looking for the best people to apply. It is not easy for a fresh graduate to enter the OECD, I would say, but you never know.
• S: It is really specialized work? What skills, either acquired at university or during other professional experiences, are required from economists in your division?
Van de VEN: As a junior economist in my division you would typically be involved in the processing of data from countries, member countries and the BRICS countries. We send out questionnaires, we get the results, these results have to be checked whether they are appropriate, they need to be published on our databases. Usually, it also involves contacts with the relevant countries, to discuss (some particularities of) their data. What I try to do within my division, is to bring more attractiveness to the job, in the sense that people do not only work on data the whole day but also do some analytical work on e.g. the international comparability of data and the caveats when you compare data across countries, and to do related publications in that field. That is something I am currently working on.
• PH: The OECD is located in Paris, do you work together with other statistical offices in Europe such as Eurostat or AMECO?
Van de VEN: My work is perhaps for a half about cooperation with other international organizations. One of the things we are working on is to have a better co-operation with other international organizations in, for example, the collection of national data. In this respect, a lot of OECD members are also EU members and what you now see is that Eurostat collects the data, we collect the data, and that is not very efficient. So we are now in a process where we discuss with Eurostat, looking at our procedures (what do we do, what do you do?), try to improve it and in the end Eurostat will become responsible for the data we use for the EU countries. But this involves a lot of discussion between the two organizations and involves quite a number of people working on this kind of statistics. There are also a lot of cooperation in the compilation and further development of the NA standards worldwide. For this purpose, there is a group active, the so-called “Inter Secretariat Working Group on National accounts”. It consists of five international organizations: Eurostat, OECD, World Bank, IMF and UN and they take responsibility for the standards. When you look at these standards, it is a rather complicated process to change them, because they are applied worldwide. To involve countries in the process, an advisory expert group consisting of 15 to 20 NA experts worldwide discusses all topics in detail. If they have come up with a new revised set of guidelines, then this has to put forward for agreement to the UN Statistical Commission where all countries are represented. Especially in the area of NA, you have a lot of exchanges between countries and between international organizations.
• PH: When, during your studies, did you decide to specialize in NA?
Van de VEN: For me it was already clear when I was 16 or 17 years of age that I wanted to do macroeconomics. I was during my studies very engaged and I was very interested in economic policy oriented questions like government debt, unemployment or economic growth. I was always very fascinated by the topic. During my studies I hardly heard of NA. Well, then you start to apply for a job and actually my first interest was to get at a job at the Netherlands Bureau for Economic Forecasting and modeling.
NA is a fascinating countrywide kind of book keeping system; the system really is a beauty of elegance, absolutely beautiful. In bookkeeping you have the double entry principle. Every transaction is recorded as an income/outlay with a concomitant change in financial assets/liabilities). With this identity, you can check whether data are complete and consistent. In NA, it is all quadruple entry, because every transaction involves always two parties: if you receive interest, somebody else is paying the interest. So, every transaction is recorded four times. However, I must say that while I was studying, the famous equation Y=C+I+(X-M) was something totally abstract for me, but when you enter into NA it becomes very concrete because it is that equation which we use the whole time when we process data. It is the basic equation of supply equals demand and what we do is trying to arrive at consistent data in which: Domestic Production (P) + Imports (I) = Intermediary consumption (IC) + Exports (E) + Investments (I) + Consumption (C). We use this identity all the time when we compile NA. We have statistics on how much is domestically produced, we get foreign trade statistics, we get household budget surveys which give information on the purchase of goods by households, we have enterprise statistics, etc., etc. We put all this information in the system of supply and use tables, and what you will see that it does not match and then we have to decide which data need to be adjusted. To give an example, if you look at cigarettes, then you would typically see that supply is larger than demand. Consumption of cigarettes is typically underestimated and we typically increase consumption, to arrive at consistent data. This equation is very important in NA. It's is also a way in which we can sometimes detect hidden activities.
• PH: Given your experience, would you recommend TSE students to remain generalized in the large field of economics or rather to specialize?
Van de VEN: My personal impression is that, especially when you look at people from the university, it is not that important what they exactly have studied. It is much more important that you have the impression “This is smart young lady or a smart young man.” In my field of expertise I do not expect that people know a lot about NA, but I want them to comprehend the system quickly, that they quickly get up to speed and really bring value added to the work we do. Of course it helps if they know something about consumption or exports, but they usually do not know the details. I could even take somebody who studied physics if I know that he/she is a smart and has a certain quantitative feeling. You should not be afraid of numbers and computers. It certainly helps if somebody has done macroeconomics, but it is more important to comprehend fast. Maybe this is different in more specialized fields.
• PH&S: On behalf of the TSE students, thank you for your time, it has been very insightful.
Important Note: Mr. Van de Ven is actually looking for an intern, starting June 2013 (or somewhat later) for a period of about 6 month. The main tasks will consist of providing support in the update of the OECD-publication "Understanding National Accounts" (available on the OECD-website). If interested, please contact him at Peter.VANDEVEN@oecd.org.
As mentioned previously, to do an internship at the OECD requires you to be a national of a member country of the OECD.
More information about the OECD available at the following website: