Being involved in educational research in Romania for more than a decade I often have the impression that, though many good quality research is being done in the area, those who are spending effort in this domain are somehow marginal relative to both the educationalists and the social science researchers, on the one hand, and the results of their research inform in a disappointing measure the educational policy debates, on the other. Among the many recent research that could have been valuable inputs for policy design in education one could mention Fartușnic (2014) on the issue of school finance, Surdu (2011) with a report on the issue of teachers' absenteeism in Roma schools, Mureșan (2014) on the impact of educational aspirations and achievements on demographic phenomena or Ion and Iucu (2014) who directly investigate the uses of research in the practice of teaching, to mention only a few that have drew my attention recently. Although a simple online search identified more than 100 educational research articles focusing on Romanian published between 2011 and the middle of 2015 in academic journals most of them have little or no impact on the way educational policies are devised and implemented.

However, there is in education and in educational research in Romanian much more than meets the eye and the purpose of explaining the actual state, especially considering the bleak situation of Romanian education, needs a more careful analysis.

With this paper I took advantage of the invitation from the Hungarian Educational Research Journal to elaborate on the topic of educational research in my own country following a presentation at the EERA Conference in Budapest in September 2015 to reflect a little deeper on the state of the field in my country. For this aim I have described the context of Romanian education and its post-1990 evolution, I have conceptualized education and educational research using the framework of strategic action field and explored the potential for longitudinal description of applied research in the area using as data source the research project list published by the Institute of Educational Sciences, an educational research body which is subordinate to the Romanian Ministry of Education. Concretely, I have investigated the longitudinal dynamics of educational research topics, correlating them with changes in governments and in political ideology, looking for eventual clues into what determines what topics gain the attention of those placed in the center of policy-making: the Government.

The scope of the descriptions and conclusions deduced from my subsequent analyzes are limited as are the data used to described the educational policy agenda or the state of education and policy making. Nevertheless, a discussion of Romanian educational policy cannot be started without understanding the main trends of the past in educational processes and outcomes.

In 1990 Romanian education was wearing all the features of Soviet style education: complete participation in compulsory education had been achieved, almost full literacy, large segments of post-lower secondary education were enrolled in vocational education and training designed for the necessities of the large industrial sector of the country (Florescu, 2016) while access to higher education was kept under strict control making the transition to the tertiary degrees one of the most difficult task faced by youngsters and their families looking for a socially rewarding educational destination (Reisz & Stock, 2006).

The two and a half decades that have passed since then witnessed dramatic changes in the Romanian educational landscape.

1. The virtual disappearance of vocational education and training in the initial education. Although at the beginning Romanian educational system could have been described at least partially as a dual one, in which large segments of pre-tertiary students had learned in forms of vocational training and apprenticeships gradually, during the first two decades after 1990, these forms of education designed for the necessities of specific industries lost their appeal and significance. As Romania suffered a massive deindustrialization forms of IVET became more and more obsolete, void of any relevance for the world of labor. After a constant decline in quality mainly these forms of pre-tertiary education have been abolished in 2009. In 2012 though, part of a re-industrialization push, they have been reinstated again but the enrollment figures prove a low demand.

2. Massification of tertiary education. While IVET was dying, enrollments in the tertiary education were skyrocketing. From 1998 to 2009 the number of ISCED 5 and 6 students enrolled in Romanian universities increased more than threefold, a rate of increase that has been rarely met in any other European country. According to analyzes regarding the expansion of higher education, demand for credentialized degrees in Social Sciences, Law and Business and Administration explain much of the expansion and the subsequent contraction, entailed by the saturation of demand in these domains after 2010. Percentage of young adults (25-34 years) that have graduated tertiary education reached in 2015 25,3% which is still rather far from the EU average (37,7%) but is much higher than the starting levels.

These first two dynamics highlight the transformation of Romanian socio-economic system from what Soskice and Hall (Hall & Soskice, 2001) might have called a rather coordinated market economy to a regime of liberal market economy. Compared to the situation two decades back one cannot notice that at least concerning education the features fit the models of liberal market economy sketched in the types of capitalism literature: lack of articulation of school and work including weak IVET and difficult transition from school to work (although overall unemployment is low, the rate of activity is unusually low in Romania and the transition from school to work rather difficult in Romania: for instance the employment rate of youth graduates 5 years after graduation are among the lowest in EU - around 66% in Romania and 77%, respectively in EU-27 in 2014 according to the Labor Force Survey of 2009; according to the Labor Force Survey of 2009, the length of job search for a graduate in Romania is 10 months compared to the EU 27 average of 6,5 months - see Eurostat) and large enrollment figures in tertiary education to which one can add the poignantly low levels of participation in lifelong learning (see Hall & Soskice, 2001). This is important as the public’s discontents regarding the education refer most often to the weak connects between the worlds of school and that of work.

3. Mediocre outcomes of education. Over the last decades the indicators of results and efficacy of the educational system had been rather disappointing: in all the OECD measurements of student competencies in which Romania participated the country’s results had been below the OECD average and without clear signs of improvement (OECD, 2007, 2010, 2013; Preda, 2009). Measures of educational efficacy – dropout rates, early leaving of education have been shaped during the same interval after the evolution of the public financing of education. This is linked to the following feature of Romanian education.

4. Inconsistent educational policies and weak state support for the educational system. After 1990 the major buzzword in Romanian education became `reform` which actually introduced an era of frequent changes justified by the state of the system and motivated by the visions and interests of stakeholders. Two major Reform Laws shaped the era, in 1995 and 2011 both passed by right-wing majorities that pushed for liberal reforms in education. Both were subsequently followed by the efforts of the left-wing governments to limit the effects of the Reform laws in order to guard the especially the interests of staff employed in the public educational sector. Ironically, reform in education was attempted with stingy resource provision, measured as proportions of public budget allocated to education. Though defenders of educational policies argued constantly that in nominal terms the amount of public resources used for education has increased significantly, statistic that has been even more augmented by the constant decline of the educational demography, the reality is that Romania has always been in Europe during the last two decades, among the countries which provided the least amount of funding for education as percent of GDP. There is no surprise that the outcomes of education in Romania are not among the best at the European level given the constant shifts in policy priorities and the underfunding which produces a largely unmotivated and negatively selected teaching staff.

But the inconsistencies of educational policy, and their fluctuations along government allegiances lead us notice another trait of decision making in the area, which is revealing about the place of research, being mainly driven by ideology instead of scientific evidence. Various political majorities acted upon education following the interests of their constituencies without much consideration for hard data, impact evaluations, ex-ante evaluations, randomized trials a.s.o. This points to one of the theses of my paper, which is the weak institutionalization of educational research as a strategic action field. Strategic action fields are autonomous, self-regulated, areas of contention (debates) in policy design, according to Fligstein and McAdam (2011) which is the case of the general educational policy discussion in Romania but cannot be used as a description for the educational research sphere which is not clearly institutionalized - with identity, internal rules and organizations. Being this absence of institutionalization stressed I will turn to my working hypothesis concerning the evolution of the educational research agenda in Romania, which I consider as a reflection of the policy agenda, on the one hand, and of the general neoliberal agenda that comes together with the influence of the large donor organizations like the World Bank, OECD or the European Union.

The field of policy oriented educational research is as diverse as one can assume upon a shallow analysis. However, after a more thorough assessment one can identify 5 main categories of actors doing applied research in the area:

1.      The Romanian Government through its research infrastructure, in this case the Institute for Educational Sciences. As is stated on the IES's homepage: 'it is a national research, development, innovation and training institution in the domains of education and youth, a unit that is linked to the Ministry of Education...' (author’s own translation). The Institute employs tens of researchers with expertise in education and related areas and carries on research tasks drawn by the Ministry of Education. The major international educational measurements in which Romania takes part (PISA, TIMMS etc.) are coordinated locally by the IES.

2.      International donor organizations, especially the World Bank, UNICEF and the European Commission implement their own research agenda in Romania pursuing their education policy goals. For this objective usually they contract independent foreign and indigenous experts who conduct research along the institutions' objectives. Major educational policy reforms in Romania have been pushed forward by these organizations usually relying on own needs analyses, impact assessments a.s.o. In several occasions these reforms have been accused of being too much in the direction of a neoliberal, market driven education. Such were the change to a per-capita financing system of higher education in the 90s, the social inclusion programs of the Romanian governments designed with the help of the World Bank and European Commission but also the various EU programs directed towards improving access to education and improvement of nation's human resources.

3.      Several powerful NGOs active in the area of children's rights and well being, mainly international ones - like Save the Children, Open Society Institute - but local ones also, benefitting from large international donations - like Romani Criss - have been pushing for policy changes in various area employing among other things research results. The welfare of children and youth from some disadvantaged groups - those with disabilities, Roma, children of transnational families - largely overlooked in the mainstream agenda is promoted by these organizations.

4.      Romanian universities and the institutes affiliated to the Romanian Academy of Sciences follow their own researches on topics which are most often less related to the issues of immediacy pursued by the other actors. Often the research agenda is one of catching up with the international research topics and paradigms. Academic research in the area is fruitful sometimes by giving independent input to decision makers but also by generating a pool of independent experts that can cooperate easily with the other actors.

5.      Beyond the above mentioned four categories we can pick a fifth one, that of local and international consultancy which provides the research skills and expertise needed by some of the main educational policy actors to design programs and policies via baseline surveys a.s.o.

Among the five classes of actors relationships are complex and not seldom competitive and/or even load with conflicts. The government has to face often the pressures towards policy changes from both the international donors and the large NGOs which, on their behalf have also a strenuous relationship, marked every now and then by alliances followed by separations. On the other hand it may happen quite frequently that academics raise critical voices against current policies, both at national and international level. This is more the case the less the academic research relies on national funds but on international one obtained through researchers' own networks.

Although educational research is being made by so many actors, educational research per se has very difficult time emerging as a strategic action field in the sense of Fligstein and McAdam. We can immediately gather some clues pointing into that direction. One important indicator is the absence of a national association of educationalists in Romania not to mention one of educational researchers. Some niche organizations really exist - like that of Romanian elementary school teachers (Asociația Românã a Învãțãtorilor) or that of Romanian Pedagogists from Covasna and Harghita - these two being the most salient after a shallow web search. Even more symptomatic for the absence of self-organizing capacity of educational field in Romania is that the main educational journal of Romania - Tribuna Învãțãmântului - does not have on its web page any link to any independent educational organization but of state ones. The absence of such professional organization highlights the weak professionalization of the education jobs and stresses its' weak self-regulatory capacity, its feeble professional identity and low levels of social capital (connectedness and trust) within the professional body. There is no wonder then that compared to the situation of other European nations, Romania is represented in the EERA (European Educational Research Association) by the University of Bucharest while most of other countries have national educational associations as affiliates of EERA. Causes for this faintness are manifolds but can be easily revealed to insiders in the systems.

First, like most other policy domains, Romanian education as a field is a subordinate one meaning it has above itself more authoritative circles - that of government first, then that of political parties with their competitions, that of trade unions, of media and experts. Education is a field of policy competition on which the regulations - laws, procedures etc. - are outcomes of strategic interactions of actors from all the mentioned superordinate levels much in the manner described by Margaret Archer in her classic work on the emergence of modern educational systems in England and France (1984). Not only the policy considerations but the people inside the educational field either are wavering among conflicting powerful groups which all have terrifying leverages upon the teachers: the political apparatus has an important say in the career prospects and in the distribution of resources, trade unions can protect against administrative abuses, a.s.o. Because of these centrifugal forces the educational profession has a hard time in consolidating a unique mission, a set of values and objectify them into an organization capable of regulating the profession from inside and insulate at least partially the educators from the mentioned outside pulling forces. A similar logic applies too for the case of the little relevance of educational research in policy design: as policy design is the outcome of strategic interaction among levels of government, political parties, and other actors there is little room for evidence-based policy other than those that serve the interests of groups in the field. Proof for the opportunistic treatment of research data is the strictly interest based placement of commentators regarding the results of international competency evaluations - usually those holding the power in the field contest everything about them while their challengers use them as tools to contest the distribution of power within the education domain.

Secondly, research is not seen as a core and defining part of the educational profession. We find evidence for this in the most recent published criteria for awarding the much desired Merit Salary Increase of 25%, a prize that is delivered for 5 years to the best quarter of teachers on the basis of achievement records where research and publications play almost no role while 80% of criteria refer to „complex activities with educational and training value”^[2]. This has to do again with the centralized, highly hierarchical practice of the education jobs, in which all organizational reflection is structured in a bureaucratic manner in which there is no customary role assigned for self-regulated practices, experiments, debates, a.s.o. To make things worse, research in education is not a homogenous set of practices that would observe a common perspective on validity and relevance. While trained educators focus in their research effort almost entirely on topics of pedagogy (effectiveness of teaching methods, class organization a.s.o.) with a low level of methodological sophistication, a lot of current educational research is trans-disciplinary, with inputs from psychology, sociology or even economics and a high degree of complexity in their use of analytical instruments. Many of those who are doing high quality trans-disciplinary research in education do not see themselves as educational researchers but psychologists, sociologists etc. Such a disconnect could entail among other causes, a weak penetration of research results in the ranks of teachers, little relevance of educators' own research outputs and, subsequently little interest for research among teachers.

Although the landscape of Romanian educational research as described above seems confused one actor stands out as having a special relationship with the main policy maker in the area: the Institute of Educational Sciences. It is a research institution directly subordinated to the Ministry of Education which employs the Institute's resources for its policy objectives. Thus, a lecture of the topics covered by the Institute gives a picture of the evidence-based policy agenda of past and present governments in the field of education.

The previously described emerging status of educational research as a strategic action field in which policy research is subordinate to the strategic considerations of major stakeholders suggests for the case of Institute of Educational Sciences a couple of working hypotheses regarding its research agenda: 1) the research topics will be vary with changes in government; 2) the topics will reflect the ideological orientation of the government: while left-wing government will be more interested in topics related to the fair distribution of educational resources to students and teaching staff right wing governments will be interested more in issues connected to the narrative they neoliberal agenda.

In order to assess the evolution of the Institute of Educational Science's research agenda in connection with the contextual factors I have done a simple thematic content analysis of the data base of research topics published by the Institute on it's own webpage (at http://www.ise.ro/teme-de-cercetare). My task has been greatly eased by the fact that the research topics are published sorted by year in a standardized way, tagged with keywords and with links to the research activities' outputs.

The IES webpage contains 92 research projects covering a time span 1991-2015 (inclusive) - though there is only one project older than 1996, all the other 91 projects have been running no later than 1996.

The analytic procedure was unpretentious: I have simply coded each of the 92 units of analysis with thematic labels describing the areas/domains of education research covered by the project. I built the list of codes partially with the help of the keywords used by the IES in their database entries while others I have picked up inductively and, in a certain way using a trial-and-error approach. The final list contains 24 codes of which 2 are rather meta-codes, referring to the type of approach in the project (survey, attitude measurement, on the one hand, and theoretical, on the other) while the rest of 22 codes refer to areas of research and intervention in education. The data table that resulted from the content analysis can be downloaded from http://www.hatos.ro/home/researches .

Finally 174 codes have been attributed to the 92 research projects - an average of 1.89 per research project - which highlights the fact that most of the project can be described either by the topic and the general approach (25 projects employed a kind of survey method to reach to their results) or, usually, by two thematic codes. However, I tried to avoid multiplying excessively the labels by attempting whenever possible to identify one or maximum two main themes.

One observation has to be made concerning the time-labels of the research projects: I have coded only the year of project inception which adds a limit in the analysis as each project is counted as a one-year one, though several of them (not many actually) lasted for several years thus deflating for several years the number of ongoing research projects.

The simple line-chart of project inception by years underline the fact that research in education follows the general pattern of public investment in education and research (research is usually a subordinate department in the Romanian Ministry of Education). While the average number of projects started by year is 5 (if we exclude the year 1991) there are several years in which projects' start have flourished relative to the previous and the next periods, especially 2006-2007 (an interval in which public investment in research had been generally more generous than in the previous intervals), but also 2012 and 2014. Whereas the drop of 2013 may hide the actual implementation of projects started in 2012 the increasing slope of 2004-2007 and the subsequent downward line of 2008-2010 reflects the manner in which publicly funded research has been affected by the austerity policies of the economic downturn.

Most frequent covered topics by the IES's projects are those in area of educational system's evaluation using statistical data (among them are notorious the Reports on the State of National Educational System that have been first published in 2005), researched covering educational policies, civic and moral education, educational management and researches on youth.

Surveys, mostly quantitative have been a major tool for policy research in education. They are done on various populations, mostly on students, but on teachers, parents and members of the community as well.

One plausible hypothesis is the variation of the research interests of the Institute of Educational Sciences according to the ideological orientation of the Government. This is suggested by the rich international policy research evidence available according to which educational policy depends on the ideology of the political majority (Busemeyer, 2009; Jensen, 2011; Kempkes, 2010; Poterba, 1997; Svallfors, 1997) and by the ideologically inspired debates concerning educational policy alternatives like those concerning marketization or privatization in education. Several of the research topics in the database of the Institute of Educational Sciences are prone to such positions: access equity, participation and dropout, lifelong-learning and the gender dimension which are more of focus for leftist movements as they relate to the relationships of education with the distribution of life chances.

In order to test this working hypothesis I have analyzed the association of topics with the ideological orientation of governments in power between 1990-2015 described simply using a left-right dichotomy (pre-2000 - mostly right^[3], 2000-2004 - left, 2005-2008-right, 2009-2012 - right, 2012-2014 - left, 2015 - right) excluding the two methodological codes - of survey and of educational theory. This was done by comparing the observed frequencies of the crosstabs of government (ordered chronologically) or of the government's ideological leaning with their corresponding theoretical frequency computed using the well-known chi-square formula.

Before getting to the results it is worth mentioning that in terms of chi-square there is no overall significant association between the variables considered – governments and the frequency of specific topics. Although the use of chi-square test in this case is strictly explorative – the number of expected frequencies smaller than 5 is so large that the computed chi-square is not usable (Agresti, 1996) – it is important to notice that no clear pattern of association can be inferred from the data and at least a strictly observational basis our working hypothesis does not stay on solid grounds.

Thus grouped and assessed, against theoretical frequency of occurrence, I was able to identify trends in the evolution of the research topics on the IES's agenda. Three topics have been high on the agenda recently: curriculum development – probably in the context of the debates concerning the new national curriculum, educational management and educational policy analysis.

Recent topics (salient topics during the last two governments):

Reiterating the analysis after grouping the data according to the ideological orientation of the government in the interval one will find the same lack of consistency. There are only several topics that can be associated with the left-right dichotomy:

More interesting than the few correlations thus noticed are the inconsistencies within the same ideological category: for example, Educational Management has been more often approached during the 2009-2012 government than the previous one, also mostly liberal, of 2005-2008, the similar pattern being found for the case of access equity, participation/dropout, resources for international evaluations and others. Looking at the data the more obvious pattern seems to be that of radical shift in agenda from one government to the next one regardless of the placement on the left-right scale (this appears more clearly comparing the 2005-2008 governments’ agenda with that of the 2009-2012 one and also this one with the next one). The most coherent agenda seems to be, in my opinion, the one of corresponding to the 2009-2012 government (center-right, Christian-Democrat in this case) which suggests a concern with effectiveness and equity. Moreover, one can enlarge the analysis by inquiring the impact of the agenda of superordinate international organizations (the European Commission especially) through processes of isomorphic processes (see DiMaggio & Powell, 1983).

Before jumping to conclusions I have to highlight that the results of the content analysis describe only in an incomplete way the policy agenda of the considered interval. Given the weak institutionalization of educational research and of evidence-based policy making, one could infer ironically that researches have been commissioned more in areas which were not of upmost interest for the government. Moreover, a better understanding of the research activities motivated by the policy interests of the government should impose differentiating between research projects which were done without policy design motivation: as a research institute the IES is eligible for most of the research financing available at home or in Europe and, moreover, is an interesting international research partner. Thus, the list of topics uncovered by our content analysis does not reveal merely what had been the concerns of the governments in matters of education at certain points of time but also the topics high on the international educational agenda to which the IES is connected in various ways especially diffusion through scholarly networks.

My aim in these pages has been to describe the Romanian educational agenda of the last 25 years and to help the understanding of its dynamics in connection with the evolution of the Romanian education. In order to achieved this I started with a synthesis of the post-1990 evolutions in Romanian education in which I highlighted the transformation of the socio-economic scaffolding from a version of coordinated market economy that have emerged after 1990 slowly into a liberal market economy, the mediocre outcomes of education and inconsistent educational policies accompanied by meagre state support for education. This is important in order to put education in the proper perspective if the aim is to make education more relevant for the world of work which cannot be done independently of considering labor laws and the structure of economy in terms of industries.

Conceptualizing Romanian education and educational research using the strategic action field framework of Fligstein and McAdam (2011) as areas of strategic interaction between competing stakeholders it appears that educational research is in an incipient stage of institutionalization as a strategic action field. There are several possible explanations: one is the subordinate place of education relative to other areas of political action and, consequently, the inconsistent use of evidence-based policy making in education where political or ideological considerations. Another correlate, if not entirely a cause, is the fact that research holds a marginal place in the educational profession.

There is, however, a way of having a glimpse into the governmental research interest in the area as the Romanian Ministry of Education has in its structure a research institute – Institute of Educational Sciences – to which it commissions most of its research tasks. Taking advantage of the fact that the IES has published a well structured and pre-coded list of research projects post-1990 I have made an analysis based on coding each project with the main area/areas of research and with the year with the aim of describing the chronological evolution of the governmental educational research agenda and of correlating the changes in agenda with the changes in political majority and of the ideological leaning of the government (left vs. right).

Although the analysis is with some significant limits there are some results that can be concluded from them:

1.      The intensity of applied research activity, measured via the number of research projects, reflects, on a year-by-year comparison, the dynamics of the public funding for education and research. This suggests not only that in periods of generous funding for education there is more money for research but also that intentions of policy change are more intense and policy makers are giving more considerations to evidence-based policy change. A vicious circle which underlies the inconsistencies of educational policy is evident here: in times of austerity policy-design is more politically or ideologically driven which is not a warrant for positive outcomes while preconditions for better policies are in place in times of state generosity.

2.      Most frequently occurring topics are: system evaluation using statistics, counselling, access equity, educational policies and civic or moral education. Most of the research documents use quantitative methods – surveying teachers, students and parents seems to be the most preferred research strategy – though there are also qualitative or mixed-methods approaches also. Comparative education reports, conceptual and theoretical approaches are rather rare if we are not considering the literature reviews that put in a conceptual framework the empirical reports. On a longitudinal perspective, recently emerging topics are curriculum development, educational management and policy analysis while past topic – no longer present on the research agenda are: teacher training, school-to-work transition, civic and moral education, counseling, lifelong-learning and adult education, school violence and the gender dimension.

3.      Comparing the research agenda across governments and political leaning does not provide clues for consistent correlations. The most striking feature of research program change is the radical shift in research priorities between successive governments: change of government entails dropping several previous topics and adopting several new ones. This is more evident as the project data base is richer, i.e. for the most recent three or four governments. The relation of these fluctuations with the effectiveness of actual policies is a matter to be discussed and analyzed yet. On the other hand the correlation of political leaning with the research topics appears to be negligible. One reason for this can be the little ideological consistency of Romanian governments and political majorities which involves that policy decisions are usually determined by strategic interactions in which ideology plays little part compared to stakeholder interests. Important processes at work though in deciding the policy research agenda appear to be isomorphic ones ignited and sustained by the contact with large international actors.

Even after discounting the limits of the results from the analyzes it appears that evidence-based policy making in the area of education is hampered by some Catch-22 situations: 1) economic hardships seem to be prone for increased politically based policy-making with much attention to constituencies’ competing interests and less regard for the collective good; 2) sustainability of policies is obstructed by the tendencies of governments to abandon previous policies and promote new ones of their own. Making educational policies more evidence driven and more sustainable cannot be achieved if educational research is not better institutionalized as a strategic action field in itself which means at least the following: a) recognizing educational research as a specific domain, officially, for instance, through a professional organization set up via the collective action of those working in the field; b) putting research in a more important place in the training and practice of teachers, even through career incentives, on the one hand, and in the design of educational policies, on the other.