1
Faculty of Economics and Business
Master of Science in Economics
Master Thesis
Topic: Field experiments on education and European Union studies and
surveys;
Study on how results from field experiments in education and studies and
surveys are being taken into account in the policy making process.
Supervisor: Submitted by: Tiago Mata Georgios Spantideas
(5855713)
2
Table of Contents
Acknowledgements………...…….…... 5 Introduction………..…...… 6 Part 1 Field experiments………81.1 The lab and the wild...9
1.2 Early twentieth century experimenting...9
1.3 The social experiments of the 60s-70s………..…....10
1.4 A recent turn ...11
1.5 Artefactual field experiments……….…....…..11
1.6 Framed field experiments...12
1.7 Natural field experiments………....….12
1.8 Do people behave in experiments as in the field? ………..…..…...13
1.9 What about the future? ...14
Part 2 Field experiments in education-Vouchers in field experiments…………....16
2.1 The Colombia experiment ………...….16
2.2 The Swiss experiment (speaking about lifelong education)...19
2.3 The contribution of field experiments to policy making...22
Part 3 European Union surveys...26
3.1 The case for EAEA...26
3.2 The nature of work done...27
3.3 The Open method of Coordination and Exchange of Good Practice...29
3.4 The Procedure………...31
3
Part 4
Shape European Union Surveys into Field Experiments?...38
4.1 European Boundaries I….………..38
4.2 European Boundaries II; from the experts’ point of view………..39
4.3 Designs on Science and Democracy ………..………40
4.4 European surveys/studies versus field experiments...41
Part 5 Conclusions...46
Appendix ...49
4 ‘Because Gods perceive future things, men what is happening now, but wise men perceive approaching things.’
Philostratus, Life of Apollonius of Tyana, VIII, 7.
5
Acknowledgments
Writing this paper has been a time and brain consuming process, since I have never before attempted anything like that. Thus, I would firstly like to thank my thesis supervisor Tiago Mata, for the inspiration, guidance, comments and advice, as well as his sense of humor. Additionally, I would like to thank John List, Janos Toth, Maria Vasileiou, Stefan Wolter and Ioanna Kastritsi Catharios, for spending some of their valuable time in order to answer my questions. Moreover, I would like to thank my parents Theodosis and Maria for the psychological support –and not only...-. Finally, special thanks to Sotiris, Jennifer, Gianni, Takis, Thanos, Nikolas and Charis for being there for me.
6
Introduction
The purpose of this paper is to present evidence, results and conclusions from certain field experiments done on lifelong education as well as early schooling. Additionally the current state of the art in European Union policies concerning the above mentioned field will be assessed. Moreover, we shall expose how scientific results from European Union studies and surveys are incorporated into the policy making process of the Union. This comparison is done in order to discuss why although field experiments are conducted inside the European Union, their results are not being taken into account in the policy making process. Therefore, the structure of the paper is as follows:
The first part will cope with aspects of the field experiment methodology that make its application important when it comes to policy outcomes. To do that, a brief historical analysis of the evolution of field experiments will be performed1 by discussing a number of core papers that are mainly concerned with the history and methodology of field experiments. The purpose of doing so is to focus on the natural field experiment methodology -for education-and its policy implications. In the second part of this paper, an investigation of the outcome of field experiments done in the past, in the above mentioned areas of education will be presented2. Thus we shall explore a couple of field experiments conducted concerning education.The use of vouchers as a subject motive will also be assessed.
Thirdly (third part of the paper), we shall discuss what the current E.U policies and goals on education are. Some of the basic policy targets of the E.U regarding education focus on early school leavers and lifelong education. This is why both domains are seriously taken into account by the Lisbon Protocol3, signed in 2000. The protocol states that there are five European Union benchmarks, set as goals for 2010 namely: a) the average rate of early school
1‘Field experiments in economics: the past, the present, and the future, by Steven D. Levitt John A. List, 2008’,
Homo economicus Evolves, by Steven D. Levitt John A. List 2008) Field Experiments, by: Glenn W. Harrison and John A. List (2004) Field Experiments in Economics* David H. Reiley and John A. List (2007).
2
‘Money Matters: Evidence from a Large-Scale Randomized Field Experiment with Vouchers for Adult Training’ (Wolter, Messer, 2009). Additionally, ‘Vouchers for private schooling in Colombia: evidence from a randomized natural experiment’ (Joshua D. Angrist, Eric Bettinger, Erik Bloom, Elizabeth King, Michael, Kremer, 2002)
3The Lisbon Strategy (the EU's overarching programme focusing on growth and jobs), underlined the importance
of EU education and training policy. It underlines that knowledge, and the innovation it sparks, are the EU's most valuable assets, particularly as global competition becomes more intense in all sectors.
7 leavers should be no more than 10%, b) the total number of graduates in maths, science and technology should increase by at least 15%, while the gender imbalance in these subjects should be reduced, c) 85% of 22 year olds should complete upper secondary education, d) the average participation of working adults population in lifelong learning should rise by at least 12.5%4.In accordance to the Lisbon Protocol of 2000, the Commission published an updated strategic framework for European cooperation in education and training in 20085. This came to enhance the previously mentioned goals, more specifically referring to the notion of lifelong learning and mobility. In other words, priority is given both to lifelong learning and learner mobility opportunities inside the European Union. Moreover, we shall examine how the European Commission takes into account evidence from surveys and studies when forming new policies and targets. The role of these studies is crucial because these are the link between scientists and policy makers. In other words, it is interesting to expose how surveys ‘transform’ into policy goals. In order to do so, the current study includes opinions both from people that work for N.G.O’s conducting surveys for the European Commission, but also experts that work for the European Union on legislation matters.Additionally, the question of how results and evidence from field experiments and studies and surveys can be transformed into policy recommendations shall be assessed from a different scope: the different ways in which science and knowledge are embodied into policies in different countries will be discussed, namely speaking about the United States of American, Britain and Germany.
The fourth part of this paper will examine how surveys and studies done on education could be shaped into field experiments. In order to evidentially support the latter argument, a couple of interviews will be evaluated6. Thus, the research question will try to answer how the adoption of the above mentioned (1st part) methodology and experimental results could contribute to the European Commission policy making process (2nd part).
4 (http://ec.europa.eu/education/lifelong-learning-policy/doc28_en.htm). 5 (http://ec.europa.eu/education/lifelong-learning-policy/doc/com865_en.pdf). 6
Due to the absence of quantitative data, certain people were interviewed. Namely, professor John List of the University of Chicago -Department of Economics-, Janos Toth, member of the European Association for the Education of Adults (EAEA) that was the main author of the paper ‘Adult education trends and issues in Europe’, and Maria Vasileiou (an expert in working groups concerning Internal and External policies of Fishery inside the European Union).
8
1. Part 1
Field experiments
When a field experiment is conducted, methods are applied in order to experimentally examine the influence and outcome of an action out in the wild, rather than inside the lab. But what is it that gave birth to the development of field experiment methodology? One could say that it was the need to study and estimate causal effects of some action. Another point of view could support the need to avoid laboratories and conduct experiments ‘out in the wild’.
Field experiments core methodology is generally to randomize subjects (or other sampling units) into treatment and control groups. After that, follows a comparison of the outcomes between these groups. This comparison should finally lead to the results/ outcome of the experiment. When it comes to the field of application of those experiments, it can be said that they are used in various domains of political, economic, psychological and social interest7. The evolution of their methodology and content is to be discussed further in the current paper. A look at the history of field experiments (mainly during the twentieth century) will be pursued. This shall be done by referring to the beginning of field experimentation and continue up to today. Moreover, a proposed taxonomy of field experiments will be exposed, following the work of Glenn Harrison and John List (2004).
Economists have been using field experiments in order to analyze certain domains/ aspects of social life like discrimination, health care programs, education, charitable fundraising, donations, information aggregation in markets and microfinance programs. As a consequence of the field experiment methodology, researchers have less control over the experiment, but the treatment followed is much closer to reality than it is in a lab experiment. By that, it is meant that the results of an intervention in the real world are examined and evaluated more effectively. This is after all the purpose of conducting a field experiment: to guarantee that the treatment could be as close to reality as possible.
.
7
It is commonly known that until recently, the majority of field experiments conducted are done in collaboration with governments, nonprofit organizations, charities or NGO’s. This can be understood if we mention that the most experiments conducted the past decades were social ones. As a result governments and NGO’s were trying to improve public welfare.
9
1.1 The lab and the wild
To begin with, let us examine how the first steps in experimental thinking and methodology where made. In his paper, Bruno Latour describes the Louis Pasteur methodology of experimental work. Pasteur, -trying to deal with the anthrax microbe problem- firstly ‘transfers himself and his laboratory into the mist of a world untouched by laboratory science’ (Latour p.144, 1983). He decides to work in a temporary laboratory next to a farm: by doing this, he achieves having big organisms next to small (micro). As a result, he learns from the field conditions by ‘translating’, in his own terms.
This translation can be understood if we consider that while the anthrax microbe out in the field is mixed with millions of other organisms, inside the laboratory it is cultivated in large colonies; in other words what is outdoors invisible due to ‘noise’, can be made ‘clearly visible to the watchful eye of the scientist’ when isolated and cultivated inside the laboratory (Latour p.146, 1983). What is essential about this methodology is that it can be introduced in a plethora of economic and other phenomena: take them inside the laboratory and study them in order to discover causal relationships.
The social interest though is big, due to the above mentioned ‘translation’: agricultural societies were interested to solve their anthrax problem through Pasteur’s laboratory. It can be said that Pasteur inserted the anthrax problem in his laboratory, for all to see. After that step, cultivating microbes in the lab becomes a reality. ‘Manipulating new objects and acquiring new skills’ (Latour p.148, 1983) results finally in Pasteur’s successful vaccine, which transforms the French society of the time, through the displacements of a few laboratories. And indeed, the French society was modified, from inside Pasteur’s laboratory, as he ‘directly displaced some of its most important actors’ (Latour, p.156, 1983).
1.2 Early twentieth century experimenting
In the first quarter of the twentieth century, namely during the 1920s and 30s, field experiments started being conducted, mainly in the agricultural production sector. More specifically, Ronald Fisher and Jerzy Neyman used them in order to evaluate and answer economic questions concerning agricultural productivity and apply treatments. Neyman -working on agricultural experimentation- underlined the relationship between surveys and experiments. According to him, randomization was of great importance in that relationship.
10 Yet he was not the one to connect his name with the notion of randomization. Fisher used the experimental concepts of replication, blocking8 and randomization while conducting agricultural field experiments.
Many researchers support that it was him that introduced the concept of randomization. Donald Campbell and Julian Stanley (p.2, 1969) argue that Fisher’s most fundamental contribution was the concept of ‘pre-experimental equation of groups through randomization’. The crucial point of the field experiment studies lies on the methodology used: More specifically (Levitt and List, p.2, 2008), the concept of randomization as an essential element of the experimental method was initially introduced in those studies.
1.3 The social experiments of the 60s-70s
During the second half of the twentieth century, certain government agencies, NGO’s and non- profit organizations conducted large scale social experiments. By social experiment, Greenberg and Shroder (2004) define an experiment that should at least meet up with those four requirements: 1) random assignment, 2) policy intervention, 3) follow-up data collection and 4) evaluation. Those early social experiments were mainly voluntary experiments. The reason that gave birth to them and modified their design was the need to measure basic behavioural relationships and causalities.
The results were used in order to evaluate fields of social policies. It is obvious that the need to improve public welfare is and has always been crucial. As a consequence, the optimistic belief that the parameters could be used to evaluate policies that had not even been conducted emerged (Levitt and List, p.1, 2008). Those large scale social experiments were conducted all over the world: In Europe, they were mainly concerned with electricity pricing projects. In the U.S -the late 60s- they were concerned with employment benefits, electricity pricing as well as allowances for houses. The main characteristic of the above era of experimentation was the focus and testing of new programs.
8
Blocking is the arrangement of experimental units into groups (blocks) that are similar to each other. Blocking reduces known but irrelevant sources of variation between units and thus, allows greater precision when it comes to estimating the source of variation under study (http://en.wikipedia.org/wiki/Design_of_experiments).
11
1.4 A recent turn
On the early 80’s, a plethora of social experiments appeared and were conducted as well. This was the second wave of social experiments, examining changes and reforms to already existing programs. This change of concern namely to existing programs, was what made them differ from the above mentioned first wave of social experiments. It is worth stating, (Levitt and List, p.1, 2008) that while over 80% of social experiments from 1962-74 tested new programs, since 1983 only approximately 33% tested new programs.
More recently, in the last decade, the point of interest has been to link the lab with naturally-occurring data. The way to do that, according to Levitt and List is ‘by systematically relaxing the controls inherent in a laboratory experiment’ (p.2, 2008). Randomization process is applied to human subjects in order to obtain identification. Usually, subjects are unaware of the fact that they are taking part in an experiment. The incentive behind this approach is to take the control out of the laboratory and settle it outdoors to the field.
One could distinguish between basic types of field experiments that have emerged. A proposed taxonomy by Harisson and List (p.1013, 2004), states that one could tell between: a) conventional lab experiments9, b) artefactual field experiments, c) framed field experiments and d) natural field experiments.
1.5 Artefactual field experiments
Artefactual field experiments include some of the large social experiments conducted in a large scale during the twentieth century. They are basically the same as the conventional lab experiments, although a main difference is that they do not use a standard subject pool. In other words, they are really close to the traditional laboratory experiments apart from the fact that they use ‘non-standard’ subjects (non-students) or participants from the market of interest
9
The conventional lab experiments employ a standard subject pool (of students), an abstract framing and an
imposed set of rules (Harisson and List p.1013, 2004). Yet the purpose of the current paper is to focus mainly on the natural field experiments.
12 (Reiley and List p.5, 2007). To make things clear, let us briefly mention an example of an artefactual field experiment, found in Habyarimana et al, 200710: African subjects are called to open a combination lock, were the winner is the quickest one. The random manipulation in this case, is whether the person that instructs the subjects on the use of these locks is a co-ethnic or a member of another co-ethnic group.
1.6 Framed field experiments
Framed field experiments, are identical to the above mentioned artefactual field experiments, but with field context in the commodity, task, or information set that the subjects use. One thing worth mentioning is that in framed field experiments subjects realize and understand that they are taking part in an experiment. Additionally, the participants’ behavior depends on the field context and the framing of the field context. The importance of such experiments lies in the fact that ‘a myriad of factors could influence behavior and by progressing slowly towards the environment of ultimate interest one can learn about whether and to what extend such factors influence behavior in a case by case basis’ (Reiley and List, p.5-6, 2007).
On the other hand, there are serious critical points concerning the feature that in those experiments subjects know that they are being tested. One could suggest that the fact that subjects are aware of being monitored could influence their decisions. More specifically, (Levitt and List, p.23-24, 2008) the fact that participants know their behavior is being monitored should make experimentalists less keen on generalizing.
1.7 Natural field experiments
Natural field experiments are similar to the previously mentioned framed field experiments, apart from the fact that the subjects taking part into the experiment do not know that they are actually participating. Additionally they do not know that they are being randomized into treatment nor do they know that their behavior is being monitored. This approach drives us to some interesting claims:
10The experiment is briefly described in ‘The Oxford handbook of political methodology’, (2008) p. 359, by
13 Natural field experiments include and combine randomization and realism. One could say that by this way, both the experimenter’s lab and the outside real world come together. This ‘bridge’ works in such a way that ends up in avoiding problems associated with the other field experiment types- including social experiments. Moreover, natural field experiments consider the treatment itself as an experiment. Because of that, their aim is to find a naturally occurring comparison group in order to mimic the control group. They have been used in a large domain of economics including the measurement of preferences, the effect of institutions on behavior, etc.
Janet M. Box-Steffensmeier, Henry E. Brady, and David Collier (2008) argue that natural field experiments have advantages over both artefactual and framed field experiments, in terms of guaranteeing external validity (generalizability). The latter is concerned with whether effects found in an experiment can be generalized to different individuals, contexts and outcomes.
More extensively, Bruce Meyer (1995) states three threats to external validity that natural field experiments should deal with (Meyer, p.153, 1995): Firstly, it is the interaction between selection and treatment: in other words, unrepresentative responsiveness of the treated population should be avoided, as the treatment group might not be representative of a certain population, or the treatment might be different from the one the experimenter wishes to examine. Secondly, it is the interaction of setting and treatment: it could be the case that the effect of the treatment may differ across -and depending to- geographic or institutional settings. Finally, it is the interaction of history and treatment, meaning that the outcome of the treatment might differ between different time periods.
1.8 Do people behave in experiments as in the field?
Following the critical points and comments that have been raised concerning the efficacy and use of field experiments, the question whether people’s behavior is biased during those experiments is put into question. Matthias Benz and Stephan Meier (2008) argue that people’s decisions can differ not only depending to whether the experiment takes place in a lab or ‘out the field’: The decision making process could also be influenced by whether the subjects know they are taking part on an experiment or not. In their paper, insights both from a lab experiment and a natural field experiment are combined, comparing the different way individuals behave in donation experiments inside the lab and in the field.
14 As a result of the latter experiment on donations, it is found that there is some positive correlation- namely from 0.25 to 0.4- between pro-social behavior in the lab and the field11. (Benz and Meier, p.279, 2008). Outcomes of this paper are suggested by the authors as of great importance, concerning their implication on experimental economics. Let us briefly mention them: Firstly, experimental measures of pro-social behavior help researchers understand aspects of behavior in the field. In other words, we can interpret from those results how similar situations in the field should be assessed. Secondly, it is implied that the notion of person-situation (psychology debate) has to be taken into account in economics, as the latter is a social science. Under that point of view, it is made clear that it is rather hard to generalize individual’s behavior, as ‘the subject’s behavior correlates weakly between various situations’ (Benz and Meier, p.279, 2008) under no influence of whether the decision making is done inside the lab or out in the field.
1.9 What about the future?
To conclude this historical overview, let us speak about what is expected to lead when it comes to the future of field experiments. According to Levitt and List, (p.4, 2008), the essence of that new approach is the collaboration with outside private parties, in the struggle to understand economic phenomena. This is because up till now the private sector has not been actively involved by taking part in field experiments. As we have seen, –previous chapters- the majority of field experiments done in the past involved social/ state involvement.
Although it is rather complicated when it comes to examining firms, the domains of concern seem really interesting: price setting, demand and supply, profit maximization, the impact of information are only some of the aspects worth examining. As it can be understood, experiments with private entities differ from the already analyzed social experiments. Those studies’ main concern is likely to be testing economic theories and exploring how markets perform. This will be different to ‘informing narrowly defined public policy debates’ (Levitt and List, p.30, 2008).
To sum up, and concerning the future of the application of field experiments, let us consider the view of List (List, p. 50, Appendix): he argues that field experiments will
11
The fact that the variation is rather high though, allows us to claim that the previously mentioned correlation could be weak.
15 continue to grow despite the critics concerning practical and ethical impediments. This will happen as people ‘continue to recognize situations that permit a glimpse into interesting economic phenomena’. As a consequence of this opinion, the estimation that field experiments will be a necessary instrument for policy process doesn’t seem excessive. In order to evaluate how field experiments influence policy making, the next chapter includes evidence from two field experiments conducted in the past, concerning early schooling and lifelong learning.
16
2. Part 2
Field experiments in education-Vouchers in field experiments
We have so far attempted a brief historical review of the development of field experiments and their methodology. In order to shed more light concerning their application, two papers concerning field experiments conducted in the past shall be assessed and discussed. Those papers were selected in order to a) expose the natural field experiment methodology and results and b) underline the role of vouchers when conducting field experiments.
2.1 The Colombia experiment
As mentioned in the introductory section, we shall explore a couple of field experiments conducted in the domain of education. ‘Vouchers for private schooling in Colombia, evidence from a randomized natural field experiment’ (2002), uses an experimental research design in order to compare educational and other outcomes of lottery winners and losers. As the authors mention, (Angrist, Bettinger, Bloom, King, Kremer, p.1, 2002) the estimates provide evidence on program results on subjects that took part in the experiment. Those are similar to the ones coming from a randomized trial. This derives from the –above mentioned- consideration of the treatment itself as an experiment.
PACES (1991) was an educational program held in Colombia, providing more than 125.000 students with vouchers. In other words, the program was born as an effort to expand school capacity and to increase secondary school enrollment rates. The vouchers used, covered approximately half of the cost of studying in a private secondary school and could be renewed as long as students maintained a certain level of satisfactory performance (Angrist, Bettinger, Bloom, King, Kremer, p.2, 2002). As a result of that, certain criteria had to be applied namely that: 1) students failing a grade had to be dropped from the PACES program, 2) the voucher could easily be transferred from one program participant to another 3) cities could decide about the number of vouchers and 4) when demand for vouchers exceeded supply, cities could use lotteries to allocate the vouchers amongst the applicants. The estimates of lottery effects were based upon a regression model12.
12
The regression model used for the survey was: , where: Yic is the dependent variable for child I from application cohort c (defined by city and year), Xi’ is a vector of individual and survey
17 There were observed no significant differences in enrollment rates: the majority of PACE applicants attended a private school in the 6th grade and most finished 6th grade, whether or not they won a voucher (Angrist, Bettinger, Bloom, King, Kremer, p.10, 2002). On the other hand, lottery winners were 6-7 % more likely to have started 6th grade in a private school. The decision to attend a public or private school was influenced significally by winning a voucher, while the decision whether to attend school was not. As a result PACES could ‘open’ places in public schools.
Lottery winners enjoyed more schooling than losers. Additionally, winners were less likely to repeat grades because a) they put more effort in order to avoid failing a grade and b) there was also a reduction in promotion standards for winners. No statistically significant effect on enrollment was observed. There was a moderately larger effect for girls, although the patterns of sex differences in the effects on private school enrollment are not clear cut: for example, male lottery winners attended an insignificant 0.12 more years schooling while girls obtained a significant 0.14 years more.
Testing
After the above mentioned procedure, the authors conducted a test, selecting neighborhoods that had relatively large numbers of winners and losers (Angrist, Bettinger, Bloom, King, Kremer, p.12-15, 2002).
Children from three Bogota neighborhoods – 1995 applicants participated in the test. In order to encourage participation, prizes were offered like refreshment, covering of the travel costs, and bicycles. The test consisted of multiple-choice test in math, reading and writing. The overall response rate was 60%, a number that shows no statistically significant difference in response rates between winners and losers. Winners scored over 0.2 standard deviations more and also had higher scores in all subsets, and there was a stronger effect on girls. Moreover, teen marriage and cohabitation were reduced for lottery winners. Winners were also less likely to be working, for example they worked 1.2 fewer hours per week (in the Bogota sample). Again this effect was significally larger for girls. But, the reduction in work
characteristics like age, sex, and whether the survey was telephone or face to face Ζi is an indicator for whether child I won the voucher lottery and δc is an applicant cohort effect to control for the fact that the probability of winning varied by city and year.
18 could be due to income effects for the household or increased incentives to spend more time studying so as to avoid failing a grade13.
Results
Winning the lottery induced households to devote more (time and money) to education (reduction of work). The higher fees paid by the winners are mainly a result of the high probability of attending private school. Additionally, those who would have attended a private school anyway, switched to a more expensive one. Winners received an average $74 more in scholarship aid than losers.
Moreover, winning households spent about $52 more on school fees this amount consists of the effects of increased private school enrollment, and a change to more expensive private schools by winners 14. Research showed that the main part of the change in households` expenditure is a result of the increased number of private school enrollment. Another potential source of the increase could be price discrimination but there is little evidence supporting the latter.
To sum up, let us mention the results of this survey: Firstly, winners did benefit from higher educational attainment, higher test scores and lower probability of teen marriage or cohabitation. Secondly, the vouchers had a stronger effect on girls in many occasions as mentioned in previous slides. Finally, (Angrist, Bettinger, Bloom, King, Kremer, p.26-27, 2002) programs like PACES can be a cost efficient way to increase educational attainment and academic achievement, especially in developing countries.
13
The authors used OLS and 2SLS estimation methods. (Angrist Bettinger Bloom King Kremer, p.15-17, 2002). The effects from 2SLS estimation were larger than those from the OLS based on the voucher status since winning the lottery is not perfectly correlated with receiving a scholarship. As a result, it seems that the 2SLS estimates are likely to be more useful for predicting an impact of scholarship programs.
14
The results concerning government budget and society could be mentioned additionally: the total social extra contribution per pupil: $24(government) + $19(households) = $43 (Angrist Bettinger Bloom King Kremer p.25, 2002).
19
2.2 The Swiss experiment (speaking about lifelong education)
Stefan Wolter and Dolores Messer in their paper (Money matters: evidence from a large scale randomized field experiment with vouchers for adult training, 2009) underline the importance of life learning and continuous education. This is why they conduct a field experiment in order to examine a) if it is basically monetary reasons (costs etc) that prevent people from participating in continuing education and b) whether increasing the participation in such educational programs could result in higher economic as well as social attainments (Wolter and Messer, p. 1-2, 2009). Additionally, one of the main points one reaches by assessing this paper, is why according to them ‘money matters’. In other words, the use of vouchers for participation in lifelong education programs is discussed. Because of the fact that the authors’ main objective was to evaluate the impact of vouchers on participation in adult education, it was necessary to design a field experiment which was to take place in the subjects’ natural environment.
As the authors mention, the notion of life learning plays a vital role in E.U policies and goals concerning education. No need to repeat that- as mentioned above- there are specific targets set for 2010 by Lisbon Protocol (12.5% adults should be taking part in education projects). This goal- a rather optimist one- could not be fulfilled by market forces alone: on the contrary, the belief that specific governmental policies should move toward this direction becomes more and more popular (Wolter and Messer, p. 3, 2009). Under that perspective, Wolter and Messer are presenting evidence from a field experiment arguing why governments should intervene in adult education funding. Another important aspect is that their experiment is demand-oriented (Wolter and Messer, p. 1, 2009): in other words, their plan involves funding subjects rather than education providers, in an effort to enhance allocative and productive efficiency.
In this point, it is worth mentioning why conducting a field experiment concerning lifelong education is of great importance when it comes to policy making: Within the E.U. there are numerous tools of supporting adult education (more than ninety according to Wolter and Messer- vouchers being one of them). Those tools however have not been tested before in a ‘real life simulator’. As a result, they are put into test by being applied. By that, an ex-post evaluation could fail, mainly because of the absence of a control group that could be always looked back as a reference. In other words, (Wolter and Messer, p.2, 2009) by conducting this field experiment on lifelong education, the authors succeed not only in testing a hypothetical
20 future launch of such a program, but also in checking the efficiency of existing models on education.
Experimental design
Adult education vouchers were sent to 2,400 Swiss random subjects. Moreover, no public-domain information was generated at any point during the experimental period, in order to make sure that voucher recipients would not know that the dispensing of the voucher was part of an experimental study. Recipients were entitled to use the voucher for an adult education module of their choice and also, they were allowed to use the voucher to pay for an ongoing adult education module. To redeem the voucher, it was necessary to send the voucher with the course organizer’s invoice to the Federal Office for Professional Education and Technology. The latter paid out the amount to participants. For the total sample of voucher recipients (2,437), the authors set a database of background variables such as prior education/training, gender, region of residence and age.
Results
The authors state that vouchers can be a suitable and crucial instrument, as one of the main results of this field experiment outlines that financial incentives work. To be more precise, (Wolter and Messer, p.19, 2009) the vouchers raised adult education participation in the experimental group by approximately 20%. This causal effect could not be expected a priori, and therefore shows that adult education and training activities can in fact be stimulated by offering money.
Additionally, vouchers could be an important means-in terms of effectiveness- for encouraging people to participate in programs of continuing education. More specifically, the voucher system raised the adult education participation of individuals with low educational attainment and those with no history of involvement in adult education. The view that vouchers can and should be used as a significal instrument is consistent with John List’s comment: he claims (List p. 47, Appendix) that vouchers are a useful tool when conducting field experiments, as they are easy to understand, simple to randomize and also a powerful potential motivator.
There are similarities between this randomized field experiment and the Colombia natural field experiment (Angrist, Bettinger, Bloom, King, Kremer, 2002). To be more
21 specific, one could mention that in both cases, vouchers are used in order to provide the subjects with monetary incentives. In other words, although the Swiss experiment mainly focuses on continuous education and the Colombia PACES program on early schooling, vouchers are the core motivation instrument used. Thus, it is obvious that vouchers are an important tool when conducting a field experiment.
Under this perspective, let us briefly mention why -according to the authors- using vouchers instead of provider financing is more effective; namely three reasons in favour of the vouchers are stated (Wolter and Messer, p.5, 2009):
• The vouchers for education, instead of any other financing instrument, provide subjects with greater incentives.
• The direct address of the ‘target audience meriting effect’ is achieved by the use of vouchers. In other words, there is no ‘deadweight loss effect’ concerning the use of vouchers, and also maximization is achieved concerning the efficient use of resources. • The so call competition effect that works in two ways: a) the subjects that follow those
courses hold their future in their hands- in economical terms-, by motivating the providers to improve their offerings and b) the providers are also encouraged to enhance cost efficiency.
Concluding, while researching if -and to what extend- results from field experiments are being taken into account by the European Union when forming new policies (or in the policy recommendation process), Stefan Wolter (Wolter, p. 61, Appendix) supports that as far as he is concerned, and always based on his own literature surveys, ‘the European Union policy in continuous education cannot be based on experimental research, as we have not come across any randomized field experiments on this topic’.
In other words, according to Wolter no field experimental results are being consulted in European Union policy concerning continuous education. This is consistent with the current paper’s argument that despite the fact that field experiments are conducted inside the European Union from universities or/and research institutions, their results are not being incorporated into policies.
22
2.3 The contribution of field experiments to policy making
Now that a brief look has been taken in order to understand how -natural- field experiments are conducted in education and what their results are, we shall discuss how these results can contribute to policy making. In other words, we shall explore what is it that makes field experiments useful in the struggle to create policy recommendations, targets and goals that lead to policies. Moreover, and because of the fact that the aim of the current study is not to be one-dimensional –speaking only in favour of field experiments- we shall assess critical points raised concerning the use of field experiments.
In a 2007 paper, List explores the notion of field experiments and their significance-amongst others- to policy contribution (List, p. 2, 2007): Firstly he is providing the reader with an introduction to the experimental methodology mainly by reffering to the historical evolution of field experiments. Additionally, he is trying to underline the importance of field experiments a) when it comes to their methodological contribution and b) when it comes to their policy making contribution. Under both a) and b) List makes it clear-hence the title of his paper- that field experiments stand as an empirical approach that bridges laboratory data and naturally occuring data.
Moreover, there is another important reason why-and how- field experiments are significant; List underlines the methodological dual nature of field experiments, the one that combines both realism and randomization (List, p. 50, Appendix).: randomization provides us with causality, and realism provides us with generalization. In other words, randomization is the one aspect that allows us to create strong causal statements while realism makes the scientist more sure that his findings can generalize to the policy setting. Realism has to do with setting the field experiment in such ways that make it look as close as possible to a real life situation. For instance, the fact that subjects do not know that they are participating in a field experiments is something that supports the notion of realism.
An illustrative paradigm mentioned in List’s paper, that can reassure the above mentioned approach, is the following: here is the way in which Franklin Roosevelt explained the inspiration of ideas for the New Deal: ‘Practically all the things we've done in the federal government are the things Al Smith did as governor of New York.’ List uses the metaphor under which, the use of small scale field experiments is done similarly to the way a federalist system uses its component parts (List, p.9, 2007): what is introduced as ‘new’ legislation at the top level of a certain federalist system, has been sometimes experimented at lowel levels and found to be succesful.
23 In other words, new legislation at the peak of the pyramid of a system derives from succesfull experimentation done in lower levels. Conducting field experiments enables the researcher to determine whether the results could be re interpreted or ‘defined more narrowly than first believed’ (List, p.3, 2007). In other words, because of their methodological nature, field experiments allow scientists and researchers to discover causality of certain phenomena and thus redefine the initial theory, or modify it in order to -better- describe reality.
Joshua Angrist (p.202, 2004), is speaking about the ‘change in American education research’. Hence, he is supporting the arguement that field experiments in education have lately become a significant tool when it comes to policy making in the United States. An example of this change, underlining the contribution of field experiments to policy making, states that the U.S. Department of Education (D.O.E) funded about twelve experimental designs using random assignment for 2001-2002. Those field experiment studies focused mainly in pre-school and early school education.
Angrist argues that the main reasons for this change in funding priorities of the D.O.E are a) the existense of an intellectual movement that ‘cuts across social sciences’ and unites researchers and scientists, and b) the political (congressional) intention to reform American Education Research. More specifically, it is the ‘No Child Left Behind act 2001’ (NCLB), that calls for education to rely on a foundation of scientifically based research, defined as research using ‘rigorous methodological design and techniques, including control groups and wherever possible random group assignments’. In other words (Angrist, p.202, 2004), one can support the argument that the reason of this change was not only the -above mentioned- academic movement, but also the political will for scientific research that could include methodologies such as randomization.
Critisism on Field Experiment Methodology
Although the benefits of the field experiments- in terms of both methodology and results- are of great importance, the existence of certain limitations concerning their application is inevitable. At this point, let us mention certain critical points that have been raised, concerning the influence of field experiments to the process of policy making. Steven Levitt and John List briefly mention that such limitations could be (p.40, 2008): the difficulty of replication, discriminating between theories, randomization bias, cooperation and its limitations, ethical guidelines and the absence of informed consent.
24 Speaking about the replication process, it is understood that it can be difficult to rerun the initial experiment on a new pool of subjects, when it comes to natural field experiments. One could also say that that difficulty, lays on the opportunistic ‘mentality’ of certain field experiments (Levitt and List, p.41, 2008). In other words, it is difficult to replicate an experiment that requires cooperation. Another limitation mentioned above, is the difficulty to tell between different theories. This happens because of the fact that the researcher has less control than inside the laboratory. Moreover, artefactual and framed field experiments could be influenced by randomization bias; although, Levitt and List (p.42, 2008) mention that according to their intuition those randomization bias should not be of great importance.
Additionally, speaking about cooperation problems -specifically on experiments with private entities- there could be observed a focus on treatments that have low social benefit, and in the same time higher private returns to a firm. This is where the experimenter could be handicapped. Another reason of handicap could be if the firm doesn’t want to have negative findings published.
When it comes to ethical guidelines as well as the absence of informed consent, it is crucial to mention that the major document providing guidelines on research ethics is the Nuremberg Code of 194715. While this is understood in health research studies, it could be argued that in some cases in the field economic experiments domain seeking informed consent interferes with the ability to conduct the experiment. (Levitt and List, p.43-44, 2008) This is why in some cases natural field experiments could be conducted. Of course, there is no doubt that field experimenters should treat justly all experimental subjects while conducting a field experiment.
Alan Gerber and Donald Green, (2008) explore -from another point of view- why field experiments have had a low profile in politics: Initially, the view that politics is rather an observational than experimental science, seemed to prevail. This approach made it difficult for field experiments to shape politics and take part in the process of policy making. Only after the behavioural revolution of the mid twentieth century did that point of view –slightly- change: again, it was argued that although the experimental method is nearly ideal for
15
The Nuremberg Code’s core manifestation was that voluntary consent is a prerequisite when it comes to clinical research studies, where consent can be voluntary if subjects a) are physically able to provide consent, b) are free from being forced and finally c) can understand the risks and benefits involved in the experiment taking place.
25 scientific explanations, it could not be used in political science due to (Gerber and Green, p.360-362, 2008) a) practical and b) ethical reasons.
An illustrative example concerning ethical problems, explores the argument that ‘poverty causes people to commit robberies’. According to the field experimental methodology, the researcher should have to randomly assign subjects into two groups, and then measure the number of robberies committed by members of the second group before treatment. Then, he would have to ‘force’ the group to be poor, and then finally re-measure the number of robberies in order to come up with results (Gerber and Green, p.133, 2001). In this example, let us expose the above mentioned practical and ethical impediments: the practical difficulties refer to the insistence on baseline measurement, as baseline measurement is not necessary for unbiased results. On the other hand, the ethical problems arise concerning the act of making people poorer instead of better off. Both the practical as well as the ethical issues that have been exposed above, indicate difficulties that have to be dealt with, in order for the field experimental method to become an essential tool when it comes to policy making.
Another example concerning ethical problems can be found in Angrist (2004): while- as mentioned above- the D.O.E. has raised funding for the conduction of field experiments on education, there has been criticism, concerning the denial of educators to refuse participation to children that might be helped by a program or to require their participation in an unproven program that takes them away from needed instruction. Two concerns emerge out of the latter statement: a) random assignment appears to require that some children (or schools) are denied services that might be useful and b) random assignment seems to force children (or schools) to participate in interventions of ‘dubious merit’ (Angrist, p.207, 2004).
Angrist is replying to the latter, by suggesting that a) in real word, access to services is always limited and b) without random assignment, the decision of who needs to be treated is usually taken by individual or bureaucratic judgements. If one recognizes the uncertainty of these judgements, then allowing some uncertainty to enter the random treatment process is logical. In other words, if one cannot make precise judgements concerning the people that really need treatment, it seems fair to allow those with equal claims an equal chance of success. This equal chance of success is- according to Angrist-random assignment.
26
3. Part 3
European Union surveys
In this part of the paper, we shall examine the role of the surveys and studies into the process of European Union policy making. This will be done in order to understand the way in which these studies are formed, their significance to the formation of policies/ policy recommendations and their scientific methodology. The purpose of the latter is to discover whether the field experiments methodology is taken into account, or what could the potential benefits of its application be. I will argue that there is a different approach between the role of the scientist and the policy maker in the United States and European Union. This will also be discussed in detail further down (part 4)
3.1 The case for EAEA16
To begin with, let us concentrate to the policy making process and the role of the surveys in the European Union. Due to the complex structure of the Union, it is not easy to tell at first sight whether and where scientists and policy makers overlap and also to what extent. There are several institutions, organisations and NGO’s that work for the European Union. Their role is basically to conduct surveys and publish papers that the European Union takes into account. For the purposes of this paper, the role and works of one of these organisations –namely EAEA17-was examined. Janos Toth- the main author of ‘Adult education trends and issues in Europe, 2006’- was the main informant for this section. The purpose is to reveal how the surveys are conducted, what their link to the policy making process is, and if there is any similarity between the methodology used and this of field experiments.
Before exposing the process under which these surveys were conducted, the notion of tender has be identified: in the words of Toth (Toth p.51, Appendix), tender is a business oriented contract between the contract authority –which in our case is the European
16
EAEA stands for European Association for Education of Adults 17
EAEA is a nongovernmental organisation (NGO) that is independent from state authority or European authority, but on the other hand is a critical partner of the European Union.EAEA is currently one of the biggest networks of non formal adult education having member organisations in 41 countries (Appendix).
27 Commission18- and the tenderer which is the service organisations- which in our case is the EAEA. Of course we can generalize and suggest that instead of the EAEA, there could be another NGO, or institution.
There are two kinds of tenders: a) the open tender and b) the restricted tender. The open tender is a ‘public’ one, which suggests that it is published in bulletins (many times on the internet), and anyone interested can submit his proposal. The restricted tender, implies that the European Commission analyses and chooses which organisations/institutions are the most professional or experts in the field (in our case it is the education/ adult education field).
The European Commission has different Directorate Generals (D.G’s) depending to each field of interest. In the case of education it is the D.G Education and Culture that has responsibility over education and training. So the EAEA is a service provider for the D.G Education and Culture and thus, a service provider for the European Commission. Instead of the EAEA, there could be a university or a research institute. This critical service providing relationship between organisations and the European Union is strictly based on a professional and transparent basis: there are deadlines, goals that have to be reached and of course funding. In other words, one could support the argument that although the relationship between the scientists, organisations NGO’s and the European Union is not a direct one-the former do not necessarily work for the European Union -,there is certain form of cooperation, namely through the calls for tenders.
3.2 The nature of work done
It is important to identify what kind of work is done concerning surveys and studies conducted for the European Union by autonomous and independent organisations, universities and NGO’s. There is a distinction that needs to be made: On one hand, there is the public consultation process. This refers to the process when different social partners are being brought to discuss issues and hence influence the public policy. On the other hand, there is the expertise work, or scientific expertise (in other words, work done by universities or research groups). The former, is not direct politics, but rather a political process. The latter is more or
18
The European Commission-formally referred to as the Commission of the European Communities) is the executive branch of the European Union. The Commission is mainly responsible for proposing legislation, implementing decisions, upholding the Union's treaties and the general ‘running’ of the Union (the European Commission webpage, http://ec.europa.eu/index_en.htm).
28 less academic work. So to sum up there is a) social consultation and b) the scientific work. According to Janos Toth, (Toth p. 52, Appendix) it is very important to get the distinction between those two, because it is risky to mix them up. Moreover, Toth underlines that the work done for the publication of ‘Adult education trends and issues in Europe, 2006’ is closer to be regarded as academic, although it embraces elements from the public consultation. In other words these kinds of studies are basically scientific work, but at the same time a political process.
The key element, whether the nature of the work is scientific research or public consultation are the notions of consensus and flexibility. The notion of consensus can be seen as a process of synthesis, not only between the different people conducting the surveys, but also between those and the policy makers: during the process of conducting the survey there were debates and also conflicting differences of opinion. At this case it is under the responsibility of the decision makers (in our case the European Commission), what to do with these findings. The key point in the words of Toth is that (Toth p. 55, Appendix) ‘as far as we can influence them, they can influence the academics too. And this is really healthy in a way, isn’t it?’
One part needs to consider the opinion of the other and fall back-when necessary-, and also one has to be flexible enough to change his opinion or modify it in a certain way. Another crucial element is the existence of common objectives: In the case of the study of EAEA conducted for the European Union, the main objective was how one could make more transparent the importance of the adult education, how one could emphasize on the field and encourage the policy makers to support more the development of this certain field. More extensively, the objective was to fulfil the gap of a certain field and try to give arguments and orientation to the policy makers including the E.U Commission in order to make further steps and efforts in this field. One could suggest that this ‘mission’ unites the tenders (in our case EAEA) with the tendering authority. So, the existence of common objectives gives the ability to modify the focus and emphasise in one matter more than another one and so on.
It is important to refer to this process as a procedure that can shape the image and popular understanding of scientists as -non making mistakes- geniuses. Toth emphasises that scientists and academic people are also ‘ordinary people that can make mistakes, can be wrong, that do not have the full picture in their mind, they have to learn, they have to change and develop their mind all the time’ (Toth p. 55, Appendix).
29
3.3 The Open method of Coordination and Exchange of Good Practice
At this point we could state the main method by which the European Union is trying to achieve coordination: Namely, the Open Method of Coordination (O.P.C) and Exchange of good/ best practice. The O.P.C is one significal tool that helps to enhance cooperation. As the European Union is not an individual state, rules cannot be directly imposed to the member states, because each one of them retains its autonomy, individuality and freedom. In other words no rules and commands can be imposed. On the other hand, there are flexible regulations and recommendations which could be voluntarily followed by member states. This is so because past experience has shown that there are many actions or activities which are competences of member countries. Thus, the European Union can only assist or support these processes.
One interesting aspect of the O.P.C. is peer learning19. Peer learning-specifically in the field of education- is an effective way to bypass the so called ‘Brussels bureaucracy’. An illustrative example given by Toth, states that one country member can learn from another one without European Union ‘interfering’ as an authority (Toth p. 53-55, Appendix). Moreover, an example concerning the exchange of good/best practice states that the development of the student-centered teaching method in a school of a member state can be explained to any other secondary school in other countries. As a result, any innovative method can be copied, developed, used and improved. An explanatory example can be depicted below (Scheme I):
19
Peer-to-peer learning involves individuals, organisations, institutions etc to share knowledge and experience with one another. It can happen through existing social networks or facilitated peer learning opportunities. Peer-to-peer learning features: a)Spreading information through formal or informal social networks b)I two-way (or more) communication, c)the recognition that every participant can be a teacher and a learner and d) Can occur through either an ongoing forum or one-time exchange.
30
Scheme I: Open Method of Coordination- peer learning example. Source: Georgios Spantideas
Speaking more about cooperation programs we could mention Leonardo Da Vinci, Comenius and Erasmus. Those programs are funded by the European Union. Because of this, there always exist certain conditions for the funding: Firstly, the sustainability of the program after the project has ended; secondly, the dissemination of the results. The latter is crucial, as the importance of the results has to exceed the narrow limits of the participants, and the knowledge gained has to be spread. In other words, the results must be publically available to anyone and also this is a systematic way to explain the good practice and how it could be adopted to any other case.
In this point it is worth remembering what List states as an important aspect of field experiments when it comes to policy making contribution: new results and legislation at the peak of the pyramid of a system derives from succesfull experimentation done in lower levels (List, p.9, 2007). In Toth’s words, O.P.C and the cooperation methods used between member states of the European Union is an efficient way to put together the experiments, to systematize them and to offer publically to anyone else to use (Toth, p.53-54, Appendix).
The comparison- according to my opinion- leads to the argument that in both cases, outcomes from experimentation in low levels can be put together, systematized and finally lead to new results. In other words, there is a parallel way in which both surveys (through O.P.C) in the European Union and field experiments (through small scale experimentation) in the United States contribute to policy making. This comparison can be made obvious if someone takes a look at the following Scheme (Scheme II):
31
Scheme II: successful experimentation in low levels leads to new high level legislation results both in the E.U and the U.S. Source: Georgios Spantideas
The notion of funding is an ambiguous matter for discussion. As it has been previously demonstrated, certain field experiments conducted in the domain of education use vouchers as a monetary reward for the subjects. According to List (List, p.49-50, Appendix) this is the case because vouchers are easy to understand, simple to randomize, and are potentially powerful motivators. Thus, they certainly are a useful tool when conducting field experiments, as they mainly motivate subjects to act as in real life.
Toth suggests that although the voucher system is a potentially effective way to fund education, it is not used in European Union tenders, applications or exchange programs. But although the vouchers are not being used while surveys and studies are conducted for the European Union, there is usage of funding as a means to enhance participation and cooperation (for example tuition fees paid for students that participate in Erasmus).
3.4 The Procedure
In order to understand how surveys and studies conducted for the European Commission are shaped into policy recommendations, let us take an overall look at the procedure: The service contract signed by the tenderer and the tendering authority suggests a number of rules, norms and steps. It is a complicated process starting after the project/tender (in our case the approval of the EAEA to conduct a study) has been selected amongst a plethora of other applications submitted. So, once the service contract has been signed, a detailed program has to be worked out, also depending on the duration of the project. Thus, a plan of action is constructed. One of the first steps is the explanatory meeting: the tender
32 authority representatives and the tenderer team meet together, and also some invited outsiders are allowed. This is consistent both with the Open Method of Coordination (O.P.C) and with the very general character of the work of the European Commission: one of the main principles is that all the twenty seven members have equal rights to comment or add something. Also, regularly for such big consultation meetings, the Commission invites the twenty seven members plus other representatives.
As it can be understood, this is a rather complex procedure, as there are different groups in such a meeting (this suggests a rather chaotic conflict of interests and opinions). Firstly, it is the tender team, the research team of experts, and on the other hand the responsible person of the D.G. Education and Culture20. Additionally, the so called social partners might be invited in such a meeting. The latter are mostly independent observers, trade unions etc. The next step is the approval of the action plan by those groups. Later on, there follow other meetings in which drafts have to be prepared concerning progress of work, comments etc. The outcome of these meetings is a continuous modification and change, always under the supervision of the tendering authority (in our case European Commission) which has the right to approve, disapprove and so on. After taking into account and embracing (or not) the suggestions stated above, the survey is complete. As a result, one must keep in mind that the final product is not the Commission’s opinion, it only reflects team’s opinion and that is why the group is solely responsible for it. But, Toth clarifies that ‘the final product is more or less mirroring the central point of the Commission’ (Toth, p. 56, Appendix).
Once the survey is completed, the Commission embraces the results and publishes the Communication. The Communication is policy recommendation, in our case on the field of adult education. To make things clear, let us mention the two kinds of documents issued on adult education. One is the communication number six hundred fourteen (614); the latter states that ‘it is never too late to learn’21. This is a short policy recommendation for the member states of the European Union. Additionally, there is another one- communication number five hundred fifty eight (558) on ‘the action plan that has to be followed’22. The latter
20
Besides the relevant D.G, there can also participate other ‘sister ‘D.G.’s for example in this case the D.G. Employment or others. Generally, as one can understand the D.G’s that cope with a certain project are the ones relevant to the field of the project.
21
Communication from the Commission Adult learning: It is never too late to learn Brussels (2006). 22
33 is not an obligatory rule for the member states, but rather a strong recommendation to follow as a principle.
In the above scheme (Scheme III), an effort has been made to describe what was said in an easy- to -understand relations diagram. The diagram demonstrates briefly that: 1) the European Union calls for tenders. Thus, N.G.O’s, universities or research institutes respond to this by conducting and then submitting surveys/studies (those are referred to as proposals). 2) The institute that fulfils E.U’s criteria takes the job and forwards the proposal to the Commission. 3) The latter discusses the proposal with a number of work groups, social partners, legislation experts and finally approves it. 4) After that, the Communication is forwarded to the Council of Ministers which either rejects or accepts it.
3.5 The legislative role of the experts
Having described the procedure by which surveys and studies are shaped into Communication, let us now consider the role of experts that work for the European Union. The purpose of doing so is to reveal firstly to what extent studies and surveys are being taken into account when it comes to legislation matters. In other words, the goal is to understand
34 how the results from those studies transform into policy recommendations. Additionally, the role of scientists into the policy making procedure is explored. As demonstrated above (chapter 3.2-3.4) we have -up till now- seen how things work from the point of view of organisations that work for the European Union mainly as service providers (in our case EAEA and their studies on adult education). Now, the point of view of experts working directly for the European Union in legislation matters shall be explored. In order to do so, the main informant of this section, is Maria Vasileiou, an expert working for Directorate General of Fishery in legislation matters (p.58-60, Appendix).
In this point, it is worth clarifying that the legislative role of the experts is similar, regardless the different Directorate General that they work for. Moreover, the process followed - between different D.G’s -when it comes to taking into account surveys and studies and shaping them into European Union policies is more or less the same. In other words, regardless working for D.G Education or D.G of Fishery (which is what this section’s informant is referring to), the process followed is equivalent23.
Each country member of the European Union is represented in the Directorate Generals of each field, by making experts form working groups and deal with common European policies. Vasileiou clarifies that, (Vasileiou p.58 Appendix) those groups basically consist of each country member’s representatives. The essence of these groups is to discuss specific targets and come up with suggestions on European Union regulations concerning policy topics. This is done because the development of a ‘Common Policy’ in each field is interrelated with effective research at a European level.
Thus, the European Union is responsible for the development of strong research and innovation points through the formulation of a complete management of each field, based on a scientific and technological development of quality. By this, it is made clear that research in each different field (in our case education) is necessary for formulating recommendations and providing the legislators with scientific advice. The last years, in order to enhance the effectiveness of the Common Fisheries Policy, the European Union has been systematically increasing the amounts of money invested for scientific research.
By assessing the Vasileiou interview (Vasileiou p. 58-59, Appendix), the link between the European Commission and institutions, N.G.O’s and universities that conduct surveys and
23
Dr Ioanna Kastritsi-Catharios is conducting surveys and studies for the European Union, working for the University of Athens. She clarifies that regardless the field differences (D.G Education, D.G Fisheries, D.G health) the processes followed when it comes to translating surveys into policy making are similar (p. 62, Appendix).
