The impact of (un)conventional expansionary monetary policy on income inequality – lessons from Japan

ABSTRACT

This paper analyzes the impact of conventional and unconventional monetary policy on income inequality in Japan, using hitherto unexplored data from the Japan Household Panel Survey. Empirical evidence shows that expansionary monetary policy in Japan has contributed to diminishing the gender pay gap through an increase in working time of women relative to men, but also to increasing the education pay gap. These effects may have materialized via the aggregate demand channel and the labour productivity channel. In contrast, expansionary monetary policy has had no significant impact on the development of the age pay gap.

KEYWORDS:

• Japan
• low interest rate policy
• unconventional monetary policy
• monetary policy
• income inequality

JEL CLASSIFICATION:

• D31
• D63
• E52
• E58

Disclosure Statement

No potential conflict of interest was reported by the authors.

Notes

¹ M0 is defined by the Bank of Japan as banknotes in circulation, coins in circulation and current account balances (current account deposits in the Bank of Japan).

² See https://www.rbnz.govt.nz/research-and-publications/research-programme/additional-research/measures-of-the-stance-of-united-states-monetary-policy/comparison-of-international-monetary-policy-measures.

³ These include the asset price bubble of the late 1980s, the dot-com bubble or the financial market frenzy before the 2007/8 Lehman crash.

⁴ For all individuals with reported values for 2010, 2011 and 2012, the value for 2011 has been replaced by the arithmetic mean of the values of 2010 and 2012.

⁵ Over the entire panel data set, across both dimensions of time and statistical unit, 1,156 observations are lost when correcting by days worked per month, and 2,038 observations are lost when correcting by hours worked per week.

⁶ Following conventional practice in regression analysis with income data, the logarithm of IALI is used due to the skewness of the distribution.

⁷ However, the robustness checks incorporate the raw data.

⁸ There is an interesting pattern in Table 2 and Figure A.2 that provides some evidence for the claim that the diminishing age pay gap is mostly a statistical effect stemming from the nature of the data. In every year that a new cohort is introduced (2006, 2008 and 2011) the age pay gap initially increases, and then decreases as all respondents get older. Indeed, the year 2011 is an outlier as discussed before, but if we consider only the first cohort for the year 2003, the second cohort for 2006 (without the first), and the third cohort for 2008 (without the first and the second), we see the pay gap between individuals aged 20 to 30 and individuals aged 46 and 61 increase. It increases by 5.7% from 2003 to 2006 and then deceases slightly by 1.8% between 2006 and 2008. This adds up to an overall increase of about 3.9% between 2003 and 2008. Hence, when we stop following the same individuals over time, the diminishing age pay gap disappears.

⁹ Another contributing factor might be Japan’s restrictive immigration policy. We thank an anonymous referee for pointing this out. In countries with more immigration the effect might indeed change, depending on socio-economic features of the migrating population, such as age, education, or language. However, this question lies outside the scope of this paper and might be a fruitful route for future research.

¹⁰ The first-order difference of the short-term shadow rate remains in the reduced model even though it is not statistically significant. It is kept as its interaction terms with both sex and education are significant, following standard econometric practice.

¹¹ This value is given by the ratio of the estimated coefficient of the interaction term (0.033) and the estimated coefficient of the dummy variable for sex (0.879).

¹² For the education pay gap the p-value would be 0.16 and for the gender pay gap it would be 0.22.