In 2012, the Bank of Japan (BoJ), launched one of the largest monetary policy experiments in the history of the 21st century. Whilst in the years since the GFC, many countries have engaged in the kind of bond buy-back programs now referred to as quantitative easing (QE), Japan’s is unprecedented in scale with the BoJ estimated to own 80% of all Japanese government bonds by the end of 2018.
First a quick note on how these QE programs (supposedly) work. The central bank buys safe assets from firms (mostly government bonds) and pays for them using new reserves. The logic here is that these firms won’t be content simply to sit on these new liquid assets that earn them no return and will instead spend the money on riskier assets like securities – what proponents of QE call a portfolio balancing mechanism. This boost to the stock market will increase the value of household portfolios making these households more confident to consume via a wealth effect. With household confidence up aggregate demand rises and voila! Economy saved.
Is this what happened in Japan? In a word, no. While happy to sell their government bonds, Japanese companies have also been perfectly content to sit on enormous stockpiles of liquid assets, with the amount of retained earnings held as cash and deposits by Japanese companies reaching a record 246 trillion yen (2.2 trillion USD) by 2015. My thesis aimed to investigate why this was happening. Why has the world’s most comprehensive QE program had almost the exact opposite effect on firm behavior that its designers envisioned? In order to answer this question, I drew from two distinct heterodox approaches in an attempt to synthesize an answer that addressed both QE as an abstract macroeconomic model as well as its context specific implementation in Japan.
First, in order to critique the QE transmission mechanism as a macroeconomic abstraction, I drew on Post-Keynesian concepts of liquidity preference and endogenous money. The real issue with QE at the conceptual level is that it clings to the old monetarist assumption of an all-powerful central bank that controls the money supply. The more money the central banks pumps into the economy, so the logic goes, the greater the bank lending multiplier and therefore the greater risk of inflation. The only difference with QE is one of desired outcomes, while Friedman and the original monetarist thinkers feared inflation, this is exactly what the designers of QE want.
However, for Post-Keynesians the logic of money’s creation is reversed. It is the demand for credit within the economy that creates money. This is because as long as banks are willing to lend at or above the benchmark rate, they can meet demand for money with new lines of credit, which will be automatically accommodated by the creation of new central bank reserves later. The consequence of this is that pumping the economy full of new money will not incentivize new lending because banks have already made all the loans they are going to make in the current economic climate. While Post-Keynesian authors usually apply this logic to banks, the same basic principle holds for firms. Pumping firms with liquid assets will not provide them incentive to invest because they could already have been investing using credit financing. Firms make investment decisions based on their confidence in the economic climate, not how liquid their balance sheets are.
If this endogenous money framework renders null the basic logic of QE, why did I dig any further into the specific case of Japan? This is because I was still curious as to why this corporate surplus hoarding behavior was so much worse in Japan than in it has been in other high-income countries that have implemented QE programs. Because Japanese firms are the world champions of liquidity preference it seemed logical that the root of this behavior lay specifically in the structure of the Japanese economy. And so, I turned to the capitalist developmental state literature that first came to prominence in Chalmers Johnson’s 1982 book: MITI and the Japanese Miracle: The Growth of Industrial Policy 1925-1975.
Johnson’s basic supposition in this literature is that Japan’s economic boom was due to a policy of co-constitutive economic rule between the keiretsu corporate conglomerates, and a powerful state bureaucracy who directed industrial policy while protecting domestic industry. However, it was not a theory of the state that I drew from this literature but rather a Japan specific theory of the firm. My argument here was that Japan’s unique growth path has bequeathed to Japanese firms three features that distinguish them from their western counterparts:
- Japanese firms adopted a long-term orientation that prioritized internal capital investments, building liquidity buffers and prioritizing market share over actual profitability. This allowed the Japanese economy as whole to experience rapid-growth even if individual firms were losing money.
- Management in Japanese firms came to operate with a high level of autonomy from shareholders, allowing firms to pursue this long-term orientation without pressure to increase short-term profitability.
- The practice of lifetime, sometimes multi-generational employment within a single firm, coupled with a corporatist welfare system put firms into the role of primary social welfare providers, somewhat akin to the role played directly by the state in western economies.
While these features were essential components of Japan’s post-war growth model, since the economic crash of early 1990s each of these features have contributed to the institutionalization of a high level of liquidity preference in Japan. The long-term strategy of prioritizing internal spending meant Japanese firms had invested in excess productive capacity that domestic demand could no longer support, inhibiting their ability to renew profitability. This dynamic was exacerbated by an independent managerial culture who, insulated from the demands of shareholders, could instead prioritise paying down debts. This cautious attitude was in turn reinforced by the fact that Japanese firms also played the role of welfare providers and (perhaps luckily) could not restore profitability by reducing workforces or cutting wages, as the nation relied on high levels of employment to compensate for a relatively bare bones welfare state. This institutionalized corporate caution, stemming from the structural features of Japanese economic development, I tentatively named path-dependent liquidity preference.
I didn’t propose this idea of path-dependent liquidity preference in order to take part in the time honoured academic tradition of making up words. Rather it is because my passion lies in the intersection between macroeconomic theory and policy implementation, specifically the potential space this area opens to combine Post-Keynesian macroeconomic theory with the context specific method used by the institutionalist and developmental schools. I believe the case of QE in contemporary Japan offers a unique opportunity for this kind of theoretical synthesis.