1. Investment in knowledge

Expenditure on tertiary education and vocational programmes, 2014
As a percentage of GDP

Source: OECD based on OECD (2017), Education at a Glance 2017: OECD Indicators, OECD Publishing, Paris. See chapter notes.


Did you know?

Over the last decade, OECD countries increased expenditure on tertiary education by about 9% and expenditure on R&D by 11%.

Investment in human capital is key for innovation, technological development and long-term growth. Such investment can take the form of expenditure on education – notably tertiary and vocational training, expenditure on research and development (R&D), and investment in enabling technologies such as information and communication technologies (ICT).

Over a decade, almost all OECD countries increased their investment in tertiary education, reaching an average of 1.5% of GDP in 2014. Expenditure on vocational training increases the average spending in post-secondary education by one-third. Spending on tertiary education grew most in Estonia (67%), followed by six other OECD countries, including the Czech Republic, Australia, Mexico (by about 20% or more).

R&D expenditure also increased in almost all countries, with the Slovak Republic experiencing the most growth (140%) between 2005-15. In 2015, the R&D-to-GDP ratio ranged between 4.25% in Israel and 0.38% in Chile. There is also substantial heterogeneity in the share of “research” compared to “experimental development”: in 2015, research comprised less than a quarter of total R&D expenditure in Israel, but over 70% in a number of countries including Latvia and the Netherlands.

From 2005 to 2015, OECD investment in ICT assets remained unchanged at 2.3% of GDP, on average. This might be explained in part by the substitution between capital investment and purchases of ICT services including increased penetration of cloud-based services, and the rapid decline in prices for ICT equipment. Interestingly, in 2015 the proportion of ICT investment accounted for by computer software and databases ranged from about 40% in Estonia, Latvia and Norway to 80% or more in France and the Netherlands.


Expenditure on education measures spending on educational institutions by governments, enterprises and private individuals.

Vocational training refers to upper secondary education level programmes granting labour-market relevant vocational qualifications to participants for direct entry into specific occupations.

Gross domestic expenditure on R&D (GERD), totals all expenditure on inputs used in performing R&D. Research is original investigation undertaken to acquire new knowledge; Experimental development builds upon research to produce new or improved products or processes.

ICT investment refers to gross fixed capital formation (GFCF) of “information and communication equipment” and “computer software and databases”, as defined by the System of National Accounts 2008 (SNA08).

Gross domestic expenditure on R&D, by type of R&D, 2015
As a percentage of GDP

Source: OECD, Main Science and Technology Indicators Database, http://oe.cd/msti, July 2017, and Research and Development Statistics Database, http://oe.cd/rds, June 2017. StatLink contains more data. See chapter notes.


ICT investment, by asset, 2015
As a percentage of GDP

Source: OECD, Annual National Accounts Database, http://www.oecd.org/std/na, Eurostat and national sources, July 2017. StatLink contains more data. See chapter notes.



Spending on education is shaped by factors such as age structure of the population, enrolment rates, and nature of demand for skills and teachers’ salaries. Data on R&D expenditure are primarily collected through surveys of R&D performing organisations. Expenditure is identified as relating to (basic or applied) research or experimental development; this can be challenging in some cases – particularly for expenditure on capital inputs to R&D or certain sectors (notably higher education) – and can cause the breakdown to be unavailable in part or in full. While the measurement of physical investment (in current prices) in ICT assets such as information technology and telecommunication equipment is relatively well established, measuring software and databases is considerably more challenging. Evidence highlights significant differences in measurement approaches in the case of software (particularly own-account software). In the case of databases, the SNA08 recommends including only the costs of physical maintenance and construction of databases as produced capital, rather than the earnings potential of the data embedded in the database itself (see Ahmad and Schreyer, 2016).