Dissecting the global growth and productivity slowdown

30 October 2019

Economic growth in the developed world and CEE has been mediocre since the crisis. The explanation lies in the weakness of ICT investment and productivity.

By Robert Stehrer
photo: istock.com/ipopba

  • GDP and productivity growth has been significantly lower in the post-crisis period in the EU and the US than before 2008. Both have increased slightly in Japan, but from lower levels.
  • The main factors behind the slowdown are lower contributions to growth from total factor productivity (TFP) and information and communication (ICT) assets, both tangible and intangible.
  • Growth in the EU has been much less driven by ICT investments and R&D compared to the US and Japan
  • Traditional drivers of growth, such as TFP, skills, and investment in tangible assets, still account for the majority of growth. However, as these traditional drivers remain weak (by pre-crisis standards), the role of intangible assets (including R&D, software and databases, innovative properties and economic competencies) are becoming relatively more important.
  • Our results, and the multiple challenges facing the global economy at present, suggest that a return to something like pre-crisis rates of growth is highly unlikely in the coming years.

Growth rates in the global economy since the financial crisis have been subdued compared to the boom phase at the beginning of this century (see Figure 1). Compared to the period 2000-06 average real GDP growth slowed from more than 2% in 2000-06 to about 1.5% in 2010-17 in the EU-15. Over the same period, growth slowed from about 5% to 3% in EU-CEE, and from almost 3% to slightly above 2% in the US. Only Japan experienced an increase, from 1% to about 1.5%.

Much has already been written to try to explain this slowdown mostly emphasising macroeconomic factors as the role the role of monetary and fiscal policies, or changes in the dynamics of trade and the global integration of production. To complement the analysis of the post-crisis growth slowdown from a production side perspective, wiiw has updated and revised the EU KLEMS database, a large dataset providing information on growth and productivity drivers at the total economy and detailed industry level (the database is accessible at www.euklems.eu[1]).

Figure 1 – GDP growth before and after the global financial crisis

Note: EU-15 includes AT, BE, DE, DK, EL, ES, FI, FR, IE, IT, LU, NL, PT, SE, and UK; EU-CEEC includes BG, CZ, EE, HU, LT, LV, RO, SI, and SK; Japan covers period 2010-2015. Source: EU KLEMS Release 2019.

Figure 1 breaks down the growth performance of the four regions to allow identification of three key drivers:

  1. Total factor productivity (TFP including changes in the skill, age and gender composition of labour for the sake of coverage across countries );
  2. Growth in employment measured in hours worked;
  3. Capital formation, split between tangible and intangible contributions.

In the case of capital formation, tangible contributions includes both ICT and non-ICT assets (the latter including residential buildings and construction, machinery and transport equipment and cultivated assets). Meanwhile intangible assets are divided into four groups: software and databases (SoftDB), research and development (RD), ‘other innovative properties’ (OIPP, including for example design activities), and ‘economic competencies’ (EconComp), including advertising and marketing activities, investment in organisational capital, and training following the recent literature on the importance of intangible assets in the modern era of ‘capitalism without capital’.

Slowdown in GDP growth reflects weaker contributions from TFP and capital

Figure 1 shows that a key part of the slowdown in post-crisis growth reflects the weakness of total factor productivity (TFP). We find that average annual TFP growth, including changes in labour composition, declined from slightly above 1% to 0.75% in the EU-15, and from 1% to 0.5% in the US. In EU-CEE, the slowdown was even stronger, from 4% in the pre-crisis period to 2.5% in 2010-17. Only Japan recorded higher TFP growth, from 0.5% in 2000-06 to more than 1% in the post-crisis period. It is important to note here that the causality of direction between TFP and overall growth performance might go in both directions, i.e. that the slowdown in TFP could also be the consequence, rather than the cause, of weaker overall growth. [2]

We identify a similar pattern when considering growth rates for labour productivity allowing a focus on changes in capital deepening (capital stocks relative to hours worked). Figure 2 reveals that in the period after the crisis the contribution of capital deepening strongly declined, particularly so with respect non-ICT capital.

Figure 2 – Labour productivity per hour worked growth and the role of capital deepening

Note: EU-15 includes AT, BE, DE, DK, EL, ES, FI, FR, IE, IT, LU, NL, PT, SE, and UK; EU-CEEC includes BG, CZ, EE, HU, LT, LV, RO, SI, and SK; Japan covers period 2010-2017. Source: EU KLEMS Release 2019.

The digital revolution has lost steam

Although non-ICT capital accounts for the majority of capital investment, it is worthwhile to zoom in on investments in ICT and the role of the various intangible assets (Figure 3). Here, we identify three important facts with respect to growth dynamics:

First, the importance of tangible (ICT hardware) and intangible (software and databases) investments has declined significantly, and contributed even partly negatively to growth in Japan. Whether this indicates that the ‘digital revolution’ (which started in the 1950s and has accelerated in the 1990s) has lost steam or whether we face a re-emergence of the ‘Solow paradox’ remains an open question.

Second, the role of ‘other’ intangible assets – R&D and other innovative properties and economic competencies [3] - proved to be more stable, with similar contributions to growth before and after the crisis. Therefore, the relative important of these intangible assets has increased.

Third, we find that ICT and intangible assets account for around 10-15% of total growth. Therefore, the ‘classical’ growth drivers (TFP and labour quality, tangible investment) still explain the lion’s share of economic activity.

From the European perspective these results also reveal the meagre role of investment in tangible and intangible ICT (and partly R&D) in the pre-crisis period. The EU has become more similar to Japan and the US because the weaker contributions to growth in the latter countries, rather than a stronger performance in Europe.[4]

Figure 3 – Contributions of ICT and intangible assets to …

…. labour productivity per hour worked growth

… GDP growth

Note: EU-15 includes AT, BE, DE, DK, EL, ES, FI, FR, IE, IT, LU, NL, PT, SE, and UK; EU-CEEC includes BG, CZ, EE, HU, LT, LV, RO, SI, and SK; Japan covers period 2010-2017. Source: EU KLEMS Release 2019.

Future outlook and policy needs

Considering the weak 2010-17 performance of the economies discussed above, we conclude that the short- to medium-term outlook is also not very rosy. The business cycle has turned downwards and the global economy faces major challenges, including a slowdown in Chinese growth (which accounts for about a third of global growth), rising trade tensions and protectionism, and general macro-economic conditions. Our research adds to existing calls for a strong policy response to the current challenges. This must include the maintenance and strengthening of global institutions and regulations. In addition, we see the need and scope for fiscal and monetary policy action to address future challenges, including environmental issues and - particularly for Europe - demographic challenges which might be combatted with higher productivity growth.


[1] The project was financed by the European Commission DG Economic and Financial Affairs (under Service Contract No. ECFIN-116-2018/SI2.784491).

[2] Note that by definition TFP contributions of TFP growth are equal to those for value added growth; for technical details see Stehrer et al. (2019). In Figure 1 the height of the bar without hours work growth indicates labour productivity growth.

[3] For measurement of these assets (which are outside the boundaries of National Accounts data) in the EU KLEMS Release 2019 see Stehrer et al. (2019); for general discussion see Haskel and Westlake (2018), and www.intan-invest.net.

[4] To a certain extent this result might be driven by measurement issues (particularly, the deflators used for ICT capital). This has to be sorted out by future research.