IS IT POSSIBLE TO ESTIMATE LABOUR PRODUCTIVITY IN THE CZECH HIGHER EDUCATION ?

This paper deals with the issues of estimation of labour productivity in the Czech higher education institutions (HEIs) and also at the Faculties of Economics of the Czech HEIs. We focus on the period between the years 2006 and 2010. At first, we analyze the influence of labour productivity on the level of average wages of academic staff in 2010. In this case, we consider that the labour productivity consist of two parts – teaching productivity (the total number of students adjusted by the coefficient of economical difficulty per academic staff) and research productivity (the total number of publication points per academic staff). Secondly, we compare the changes between teaching productivity in the period between the years 2006 and 2010 and the changes between average wages adjusted of average inflation rate at the level of HEIs and at the level of the Faculties of Economics.


Introduction
In 2009 the White Paper on Tertiary Education 1 was brought out and the discussion of tertiary education reform has significantly gained on importance since then.It brought questions of quality of higher education institutions (hereafter: HEIs) and academic staff as well.That is why we decided to present analyses dealing with labour productivity and labour costs at the Czech HEIs and at the Faculties of Economics.According to The Principles and Rules of Funding of HEIs 2 , public funds (subsidies from the Ministry of Education, Youth and Sports for educational activities of HEIs) are allocated to the level of HEIs and not to the level of faculties.Allocation of the university budget to individual faculties fully depends on autonomous decision of the Academic Senate of HEI (the Senate has to confirm the Rector's proposal of the HEI's budget 3 ).The main goal of the paper is to evaluate whether the allocation of HEIs' budgets on faculties leads to the significant relation between teaching and research performance of academic staff on one hand and average wages on the other hand at faculties related on economic branches of study.Measurement of performance and productivity in non-market industries is a very demanding issue.While the productivity in market industries can be considered as a ratio of sales (adjusted of changes in own-produced inventories) to employment, in non-market industries we cannot measure sales as an output.As a key reference to an issue of non-market-industry productivity we consider Atkinson Review (ONS, 2005); chapter 9 is devoted to education.Consequences of differences between marketindustry productivity and non-market-industry productivity including estimates of production function for non-market industries are presented by Simpson (2006).However, both Atkinson and Simpson use British data and in relation to education they take into account mainly basic and secondary education.Productivity in higher education and approaches to its measurement are defined by Gates and Stone (1997).As most important in this paper we consider terminological differences between terms efficiency and effectiveness.Jablonsky (2011) analysed the efficiency of teaching and research activities at the level of departments using DEA methodology.Huzvar and Rigova (2012) used DEA methodology for analysis of relations between academic process and funding of public HEIs.Finally, relation between productivity and policy making is introduced by Callan (2007).The aims of the article are (i) to estimate the relation between average wages and academic performances and (ii) to compare differences between changes in labour productivity and changes in labour costs (represented by the average wages) both at the Czech HEIs and at the Faculties of Economics between the years 2006 and 2010.

Material and Methods
For the analysis we use data from the Ministry of Education, Youth and.This data set includes data on average wages of academic staff, number of academic staff (MŠMT, 2012a), average number of students (MŠMT, 2012b) and the sum of the publication points (called "RIV points") using the "Methodology of Evaluation of Research Institutions Results and of Evaluation of Finished Programmes 2011" (RVVI, 2011).All the analyses are presented only on the public HEIs excluding artistic HEIs which are the outliers 4 .Colleges are not included into the analysis.Due to the lack of the dataset needed for the analysis we had to exclude the Faculty of Economics of University of South Bohemia in České Budějovice and two newest nonuniversity HEIs 5 .Firstly, we would like to find out if the average wage of the academic staff is the function of labour productivity and if there is a correlation between these variables.Labour productivity in this case consists of two self-independent parts -teaching (number of student adjusted by the coefficient of economical difficulty per academic staff) and research (RIV points per academic staff).Multiple regression and multiple correlation coefficients 6 were used for the analysis.The analysis is based on the hypothesis that changes in the dependent variable y (average wage) are caused by two independent variables x 1 and x 2 (teaching productivity and research productivity) which is presented by the formula: (1) By using the method of least squares we can estimate the multiple regression function (2) For the discussion about the relation between the variables the multiple correlation coefficient r yx1.x2... xp and the coefficient of determination R 2 were used.

4
The average coefficient of economical difficulty of the artistic HEIs reaches 5.9.It is much higher than the rest of the HEIs. 5 We do not have RIV points of the non-university HEIs .6 For more information see Hindls et al (2004).
The second part of the paper is focused on the analysis of competitiveness of the HEIs and Faculties of Economics by using the condition modified to the non-market industry where C…labour costs 7 Y…number of students After an adjustment we can state ) which could be interpreted as a requirement of slower increase of labour costs in comparison with the change of number of students.After the division of both parts of the inequation by the labour force index represented by the index of number of academic staff, we get (5) and after the algebraic adjustment (6) It means that average labour costs should increase more slowly than labour productivity 8 .Alternatively, we can consider compensation of employees as C, but in short term we can suppose the constant ratio of social contributions to wages and salaries.It implies that the inequation (4) expresses the relation between average wages and labour productivity.Since we estimate real labour productivity by using natural indicator, the average wage has to be real too.It is necessary to take into account inflation represented by consumer price index 9 .

7
Labour costs are broadly described in Jílek, Moravová (p. 129, 2007).8 The competitiveness concept is only one of the possible approaches.In the case of an increase of the quality of education (e.g.employment more qualified academic staff) the condition would be applied vice versa.9 See Hindls et al. (p. 381, 2004).

Higher Education Institutions
For the analyses we use the data on number of students of the HEIs, the coefficient of economical difficulty, number of academic staff, average wage, RIV points (see From the results and the t-statistics (presented under the regression model) we can see that both the variables are significant (5 % level of significance).There is positive correlation between average wage and teaching productivity and there is positive correlation between average wage and research productivity.
It means that increase in teaching productivity (represent by students per academic staff) causes the increase in average wage and increase in research productivity cause the increase in average wage as well.There is no multicollinearity 11 (there is weak negative correlation between teaching productivity and research productivity, which means that increase in teaching productivity cause decreasing at the research productivity).It means that the model is estimated right.It is necessary to note that this model presents only part of the average wage.
The second part of the analysis focuses on the average labour costs (represent by real average wage) and teaching productivity (measured as the ratio of students adjusted by the coefficient of the economical difficulty per academic staff) among Czech HEIs between the years 2006 and 2010.
From the results (see table 3) one can see that there are some differences in teaching productivity and average labour costs among 20 Czech HEIs.The biggest difference between teaching productivity and labour costs during the period 2006 and 2010 was achieved at the Institute of Chemical Technology Prague (ICT).The total gap in 4 years was about 19 per cent.On the other hand teaching productivity increased more quickly than labour costs especially at the University of Hradec Králové (UHK).Teaching productivity increased of 15.68 per cent in the period in question.Average labour costs at the University of Hradec Králové decreased of 12.38 per cent.When we compare this result with the first estimation presented in Fischer, Vltavská (2011) which was done for the period between the years 2004 and 2009 we can see that the biggest difference between teaching productivity and labour costs during the period in question was achieved at the University of Ostrava.The total gap in 5 years was about 31 per cent.On the other hand teaching productivity increased more quickly than labour costs especially at the University of Economics in Prague.Teaching productivity increased of 23.81 per cent in the period in question.Average labour costs at the University of Economics increased of 6.82 per cent.

Faculties of Economics
The dataset used for the analyses contain number of students of the Faculties of Economics of public HEIs, the average number of academic staff, average wage, RIV points (see Table 4 and  Table 5).First part of the analysis is devoted to the analyses of the relationship between teaching productivity (x 1 ) and research productivity (x 2 ).y = 14 085 + 508.4x 1 + 67.6x 2 (1.87) (1.24) (2.81) (8) We can conclude that the increase in the education part of teaching productivity by one causes the increase in average wage by 508.4 CZK.One point of increase in RIV points per person leads to the increase in the average wage by 67.6 CZK (both under the condition of ceteris paribus).
The values of characteristics are as follows: R 2 = 0.37, r x1x2 = -0.072,r yx1 = 0.56, r yx2 = 0.21 The t-statistics (presented under the regression model) show that variable "teaching productivity" is statistically significant at 10 % level of significance.A weak negative correlation between teaching productivity and research productivity was traced.That means that an increase in the teaching productivity caused a decrease in the research productivity represented.On the other hand, there is a positive correlation between average wage and the teaching productivity and average wage and research productivity.One can see the link between the results of the HEIs and the Faculties of Economics.There is no multicollinearity (10 % level of significance).Table 6 presents the results of productivity analysis.One can see that there are some differences in teaching productivity and average labour costs among Faculties of Economics.At almost all Faculties of Economics, increase in teaching productivity is higher than in average wages.It could be explained by three causes.Firstly, the increase in number of students recorded between 2006 and 2010 had started at the beginning of 2000s.Secondly, due to the economic recession and fiscal restrictions the total budget for public universities decreased from 2009.
Thirdly, the increase in number of students is realized mainly in economic branches of studies.But, there is a difference between individual Faculties of Economics.The highest difference between teaching productivity and real labour costs is higher than 40 percentage points (Faculty of Informatics and Statistics, University of Economics, Prague), but, on the other hand, at two faculties the decrease in productivity is higher than decrease in real labour costs.

Number of students, the average coefficient of economical difficulty, number of academic staff, RIV points, public HEIs, 2006 and 2010 10
Table1 and Table 2).
Firstly, we discussed whether average wage is the function of labour productivity represented by teaching productivity (x 1 ) and research productivity (x 2 ).y = 19 069.8 + 303.92x 1 + 76.62x 2 (4.64) (3.31 (4.29)(7)From the results we can say that increase in teaching productivity (the increase of one student per academic staff) evokes increase in average wage of 303.92 CZK.One point of increase in RIV points per person leads to increase in the average wage of