Koo–Krugman Paradox?

I’m currently reading through Richard Koo’s The Holy Grail of Macroeconomics, which is turning out to be incredibly interesting. So far, I do have some disagreements, but they’re the kind that make you think, “This is something that I need to look into” — i.e. they’re the disagreements that make you doubt yourself. For those of you who haven’t read Koo’s book, based on what I’ve read so far I do recommend it.

I’d like to share one thing in particular from Koo’s book,

[A]fter the bubble collapsed in Japan, not only were there no willing borrowers, but existing borrowers were paying down debt — and they were doing so when interest rates were at zero. Technically insolvent companies, struggling to pay down debt and repair balance sheets hit by the nationwide plunge in asset prices, were not interested in borrowing money, regardless how far the central bank lowered rates. In this environment, monetary policy by itself no longer has any effect.

— p. 29.

If we think about it in terms of supply and demand for money, Koo is saying that there is insufficient demand for money. He still advocates price stabilization, but his solution is an increase in government spending to make up the gap between demand for and supply of money.

While Koo does mention Krugman’s Japan paper, he only mentions that it credits deflation as the main culprit for Japan’s recessionary woes. He doesn’t, at least so far, comment on Krugman’s solution: targeting inflation to increase inflation expectations. Yet, there may be an important incongruity between the theories of a balance sheet recession and of forward-looking monetary policies. (Note: Koo does spend time criticizing monetary policy, but considers only conventional monetary policy, with his argument boiling down to explaining why there’s a liquidity trap.) None of this is new, or something that economists are privy to, but I’d thought it’d interest those who haven’t read Koo or followed critics of this particular facet of the balance sheet recession argument.

During the recessionary, or what the book refers to as the “yin,” phase, businesses should be fixing their balance sheets. This requires what amounts to “saving,” as firms draw from revenue flows to pay down their liabilities. (In this case, to clarify, the creditor doesn’t add to spending, because banks — the creditors — can’t find borrowers.) Krugman’s theory, in turn, seeks to increase inflation expectations, pushing the private sector to spend cash; if we see inflation as a tax on non-interest bearing liquidity, those expecting higher inflation will swap these types of assets for those with higher returns. But, this still doesn’t address the fact that what’s holding back spending is private sector deleveraging.

Some time ago Krugman suggested that Koo isn’t entirely right,

Koo’s argument is that interest rates and monetary policy don’t matter because everyone is debt-constrained. That can’t be right; if there are debtors, there must also be creditors, and the creditors must be influenced at the margin by interest rates, expected inflation, and all that.

In the very next sentence, he gets to the heart of the issue, “That said, widespread credit constraints presumably reduce the number of players who can take advantage of lower rates.” Koo argues that the creditors are banks, and these are firms that aren’t sensitive to falling interest rates and increased inflation expectations (if nobody wants to borrow, there aren’t assets banks can buy).

I wonder if it’s Koo, in any case, who pushed Krugman towards advocating fiscal policy, rather than focusing on monetary policy.

  • http://bubblesandbusts.com/ Woj

    I haven’t read Koo’s book yet, but have been equally fascinated by his ideas through various other works. Regarding the difference between Koo and Krugman on the creditor-borrower dichotomoy, I think it boils down to whether one accepts the loanable funds idea or not. Krugman, seemingly accepting the former, believes that creditors (banks) are limited by funds from the private sector and to some degree central bank reserves. In that case it would be true that a relatively large portion of the population would not be debt-constrained.

    Now consider the opposing view, that banks create deposits and are only constrained by capital requirements (to a degree) and the willingness of borrowers to accept the bank’s liabilities. Banks can therefore lend well in excess of current income and savings. In this scenario, a very significant portion of the population could become debt-constrained (with only the top X percent remaining unburdened. If this is true, debt-constraints become far more powerful and could render monetary policy ineffective even with higher inflation expectations.

    Personally I find the second example far more convincing given my readings and experience. I should also note that while demand for credit is weak, banks may also tighten credit restrictions (since there is some level of debt-to-income where a potential borrower presents an expected loss on further borrowing).

    On a slightly unrelated note, your recent work/postings have exceeded by already lofty expectations. Keep up the great work!

    • http://economicthought.net/blog JCatalan

      Something I wonder about is why Japanese banks, if they were willing to lend, didn’t use their excess reserves to buy foreign assets. During the late 90s, much of the rest of the advanced world was going through a boom, and the same is true between 2003–07. Koo talks a lot about why foreign banks didn’t come to Japan to lend, but he doesn’t really talk about Japanese banks’ involvement overseas. There are a lot of possible explanations, but it’s not something Koo discusses.

      I haven’t really figured it out yet, but I think the truth is probably somewhere “between” Koo and those who instead emphasize the banks. What I mean by that is that I think they’re approaching the same phenomenon from different angles, and that a better explanation might be achieved through some kind of synthesis.

      • http://bubblesandbusts.com/ Woj

        I’m less certain about monetary operations in Japan than the US, but here is my take given knowledge of US operations…

        Excess reserves for the entire financial system can only be extinguished by private banks in a couple manners:
        1) Banks create enough new loans such that excess reserves become required (extremely unlikely)
        2) Banks exchange reserves for currency
        Therefore, individual Japanese banks could have exchanged excess reserves for foreign assets held by other Japanese banks (since reserves must be held by national institutions…I think) but the financial system as a whole could not have done so. Considering that foreign assets were performing well, while reserves paid minimal interest, I doubt many banks cared to engage in that exchange.

        • http://economicthought.net/blog JCatalan

          Japanese banks can’t use high-powered money to buy foreign assets? But, yea, the exchange-rate mechanism is what I had in mind (Japan had a current account surplus throughout the period, IIRC).

          Edit: But, the exchange-rate mechanism might also explain why foreign banks didn’t invest in Japan. Most advanced economies at the time probably had trade deficits, so capital account surpluses.

          • Silvano

            Just two hints which could be useful::
            1. A combination of high default rates and “evergreening” of already granted loans (which is a kind of “zombiefication” of the asset side) I think it’s enough to explain why banks were reluctant to part with their own liquidity.in favor of foreign risky assets.
            2. Maybe I’m wrong but it seems to me that the Japanese banking system financed carry trade globally through short term loans, repos, etc for more than a decade lending yen to monetary & financial institutions.At least until also the Fed and the Ecb moved toward a ZIRP.

          • Silvano

            PS.: I didn’t stated two opposite things as it could seem. I meant they lent money usually avoiding currency mismatch, minimizing time mismatch and counterparty risk. Mostly short term credits to foreign financial institution, T-Bills and similar stuff.

          • http://bubblesandbusts.com/ Woj

            As I said, I could be wrong, but my understanding of modern monetary operations is that domestic banks can transfer reserves within the domestic financial system but cannot reduce excess reserves aside from transferring them into currency. So individual Japanese banks could trade reserves for foreign assets within the domestic financial system but could not trade reserves with international banks for foreign assets.

  • Pingback: Koo on Krugman | Economic Thought