Lost in translation
18 July 2017
Using a growth accounting framework, it is not difficult to identify the causes of Japan’s economic woes. In the immediate aftermath of the early-1990s bubble, the growth rate of real per capita GDP fell as the nation sought to deal with excess capacity and bad debt. This stagnation persisted into 2000s and beyond as the working-age population began to fall – even as the productivity growth rate stabilised. The prognosis for Japan has been relatively simple too: find a way to increase the workforce and growth will be restored. Unfortunately, barriers to progress have proved high; immigration is politically untenable and rapid ageing and a low birth rate a recipe for population decline. Plugging in future population estimates, former BOJ Governor Shirakawa estimated potential growth at between 0.4-0.9% in the 2010s and 0.2-0.7% in the 2020s (see Chart 7).
That narrative has changed of late. The government has launched a growth revitalisation strategy to lift longer-term growth. Judging by both official and OECD estimates of potential growth, the strategy is working. However, there are surprisingly large conflicts in these estimates that serve to undermine their value. According to the BOJ’s estimate, potential growth in the six months to end-Q1 2017 stood at 0.75%, returning to pre-crisis levels. However, the contribution of total factor productivity (TFP) to potential growth dropped sharply during this time (see Chart 8). Instead, growth in numbers employed was the main driver of the rise. The OECD finds a similar improvement in potential growth, estimating it at 0.71% in 2017. However, rising labour productivity is the primary driver here, with per capita output improving significantly in the last few years. Why the discrepancy? Partly this is a methodological issue. The OECD adopts a more conventional approach to calculating potential output, where estimates of labour, capital and TFP are derived from trend data. In contrast, the BOJ estimates the output gap first, using a labour and capital input gap. The potential growth rate is derived by computing the weighted average of average labour and capital input growth rate and the TFP trend growth rate which is estimated separately. While there is no ‘right’ way to estimate potential growth, the conflicting evidence highlights the risk of a proliferation of estimation techniques in recent years. Policymakers in Japan appear particularly prone to alternative conceptualisations. However, this can lead to accusations of a lack of transparency and a heightened risk of policy errors.
For that reason, we prefer to use a growth accounting framework when examining potential growth in Japan. However, in doing so we are willing to question assumptions that may underline component trend growth. For example, the rise in participation in recent years looks increasingly structural. A breaking down of barriers to entry and exit to the workforce, such as age and gender, are unlikely to be reversed. However, this alone is unlikely to allow Japan to fully escape the drag of its dismal demographics. To raise its potential then it would need to increase productivity. Unfortunately, the government’s reforms are tentative and unlikely to push per capita output significantly above the OECD average. In conclusion, Japan is likely to avoid the worst predictions of a post-growth economy, but policymakers may still struggle to meet their growth remits.