By definition, the rankings reflect the relative performance of countries, but in order to look at general trends over time it is necessary to look at absolute values of our various measures. It is convenient to do this by comparing median absolute values for selected variables from the 2016 ranking with those from the 2013 ranking, the first year that covered 50 countries. Since most of our measures are positively related to GDP per capita the absolute median values are best thought of as relating to countries at middle levels of income. Nevertheless, most trends are world-wide so that movements in median values are more generally applicable.
As a share of GDP, the median level of government expenditure has increased from 1.10 to 1.19 per cent of GDP but research expenditure has fallen from 0.40 to 0.35 per cent of GDP. The extra government spending seems to have gone into teaching and learning: median participation rates have increased from 64.3 to 66.9 per cent. The gap between enrolment rates and the educational qualifications of the labour force has narrowed, particularly for developing countries: the median percentage of the labour force with a tertiary qualification has increased from 20.9 to 27.5 per cent. The gender balance among staff is slowly changing: the percentage of female staff has increased from 41.1 to 43.0 per cent, at median values.
Research outcomes have improved. The median number of articles published per million of population has risen nearly 50 per cent from 1028 to 1504, although this is really over a five-year period rather than three as in the 2013 rankings we used a five-year average of publications. Similarly, median citation rates have increased by 12 per cent.
The data reflect growing internationalisation of higher education. Student movement has increased: the median percentage of international students has risen from 3.5 to 3.9 per cent. At the median value, the percentage of articles with an international co-author has increased from 38.8 to 41.5.
In our rankings we make considerable effort to recognise that national priorities in higher education vary with levels of economic development and the structure of the economy. Some of the measures used in the main U21 ranking, such as journal publications and expenditure per student, primarily reflect the priorities of the education systems of developed countries, systems which all countries may aspire to as economic development proceeds. An auxiliary ranking shows how a country is performing in relative terms given its level of income per capita. A third ranking attempts to measure productivity of the tertiary education sector given its resources and government policies. Together, the different approaches present a rounded picture of the strengths and weaknesses of each of the 50 national systems we evaluate.