Comparative Perspectives for the Middle-Income Trap
Numerically, the threshold for escaping the middle-income trap and graduating to the high-income status is defined as requiring a GNI per capita of $12,746 at 2011 prices . The number of economies that made this transition is limited . Only about one quarter of the current OECD member countries succeeded in doing so . The middle-income trap is also persistent in emerging economies since economic growth tends to slow down once the lower hanging fruits of technology transfer are harvested. In contrast, those countries that succeeded in making the transition have one important feature in common: they have excelled in developing advanced, indigenous technologies and in investing in education.
Fig. 1 Depiction of the relative standing of countries over a 50-year timeframe (redrawn for current years based on World Bank dataset . Original graph is provided for 1960 and 2008 in Ref.  )
Based on relative GNI per capita values over the course of 50 years, Fig. 1 indicates that countries such as the US, Canada, and France have sustained their economic growth. South Korea, Japan, and Ireland have overcome the middle-income trap and are now their challengers. The next group of countries is the group that remains in the middle-income trap. China, Indonesia, Mexico, and Turkey are examples of countries in this area, although this does not necessarily signify lack of progress. Based on the Global Competitiveness Report , both Mexico and Turkey are indicated as countries in 'transition' to innovation-based economies. In contrast, some of the oil-rich countries, such as Saudi Arabia, will be placed above the middle-income trap. The GNI per capita of Saudi Arabia was $26,260  (log of percentage relative to US GNI per capita in 1963 is 1.71), close to the level of South Korea.
Table 1 compares three countries that have overcome the middle-income trap to become innovation-based economies with four countries that remain in the trap, but exhibit important signs of progress. The latter also includes the efficiencybased economies of China and Indonesia. For the sampled countries, Table 1 provides the values of gross domestic expenditure on research and development (GERD) and educational investment, both given as percentages over the values of gross domestic product (GDP).
For the innovation-based economies in Table 1, the average value of GERD over GDP is 3.2 %. For South Korea, Japan, and Taiwan the values are 3.4, 3.3, and 2.9 %, respectively. The average value for educational investment as a
Table 1 Comparison of countries based on GERD and educational ınvestment over GDP (2012) [7–12]
percentage of GDP is 6.1 %. For example, South Korea invests in education at
7.6 % of its GDP, whereas Taiwan and Japan invest at 5.8 and 4.9 %, respectively. Certainly, GERD and educational investments over GDP are relatively high in these countries.
For those countries that are in the middle-income trap, the values of GERD and educational investment over GDP are comparably lower. The average values for Mexico and Turkey that are in the middle-income trap, but in the process of transition to innovation-based economies, are 0.7 % for GERD and 3.4 % for educational investments, both over GDP. The GERD of Mexico stands at 0.4 % of GDP while investment in education is relatively higher at 3.5 % of GDP. The GERD of Turkey is higher at 0.9 % of GDP and the investment in education is 3.2 % of GDP. The average values for China and Indonesia, which are in the middle-income trap as efficiency-based economies, are 1.2 % for GERD and 3.3 % for investment in education, both over GDP. Indonesia invests the same share in educational investments as China but spends less on R&D as a share of GDP. In contrast, while the value of GERD over GDP is only 0.7 % in Indonesia, it is 1.7 % in China. Even within the middle-income trap, GERD and educational investments over GDP vary for countries that are in a state of transition to innovation-based economies and efficiency-based economies.
-  Saudi Arabia is not included in Fig. 1 due to lack of data for 1963 in the World Bank dataset.