Explaining Piketty: inequality and the financial crisis

Ryan Bourne complains that my takedown of Sumner didn't actually address Sumner's main criticism of Piketty. Indeed, that is is a fair complaint. I was so busy disagreeing with Sumner that I failed to explain Piketty. So I shall remedy the oversight now.

Here's the paragraph from Piketty that Sumner critiqued:
"In my view, there is absolutely no doubt that the increase of inequality in United States contributed to the nation’s financial instability. The reason is simple: one consequence of increasing inequality was virtual stagnation of the purchasing power of the lower and middle classes in the United States, which inevitably made it more likely that modest households would take on debt, especially since unscrupulous banks and financial intermediaries, freed from regulation and eager to earn good yields on the enormous savings injected into the system by the well-to-do, offered credit on increasingly generous terms."
This paragraph does need some explanation. It is far too easy to interpret Piketty as saying that stagnant labour wages encourage low-to-middle income people to borrow. Clearly this is nonsense: under normal circumstances we would expect stagnating wages if anything to discourage borrowing, since people would delay leveraged purchases until their wages started to improve. But it's not the point. Piketty is doing macroeconomics. This is about sector balances*, not the rational behaviour of households.

When low-to-middle incomes stagnate but top incomes continue to rise, saving naturally increases. This is because people with high incomes spend a lower proportion of their incomes than poorer people do. Since saving = investment, that saving must be invested somewhere. It can be invested in assets such as property, gold, fine art and metals. Or it can be lent out, either as direct lending (perhaps via an intermediary) or in the form of securities purchases**. The balance between lending and real assets in a portfolio tends to depend on the investor's attitude to risk: a more risk-averse investor is likely to have more real assets and fewer financial ones, and the financial assets are likely to be less risky ones - perhaps good quality corporate and government bonds rather than junk bonds and equities. Investors tend to be less risk-averse in boom times and more risk-averse in downturns. During the years before the financial crisis, therefore, investors - including the earners of top incomes - tended to lend out their savings rather than buying real assets.

So far so good. We have explained why US inequality rose during this period: top incomes rose while the rest stagnated, and as high earners have a lower marginal propensity to consume, that increase translated into higher saving and therefore into rising wealth for those people. The corporate sector also ran a structural surplus during this period. There should have been a rising saving ratio.

But there wasn't. In fact the saving ratio actually fell. And this was because, as Piketty notes, the low-to-middle income earners whose wages were stagnating were borrowing heavily. Why were they doing this?

Piketty suggests that it was because of aggressive lending practices by banks. This is probably true. But actually it is beside the point. There had to be high borrowing somewhere in the US economy. With the corporate sector in surplus, either the household sector or the government - or both - was forced to borrow.

To explain this, we need to look beyond the borders of the US. At the time of the crisis the US was running a large and growing trade deficit. This was matched by large trade surpluses in other countries, principally China, Germany and Japan. As I explained in the post linked in the first line of this post, countries that run trade surpluses are net exporters of capital, and countries that run trade deficits are net importers of capital. The relationships are not necessarily bilateral: for example, country A may run a trade surplus with country B (which therefore has a trade deficit), and export capital to country C whose banks lend it on to country B to fund its imports from country A. Whatever the actual flows, however, the effect is the same: country A is a net exporter of capital because of its trade surplus, and country B is a net importer of capital because of its trade deficit. To simplify things, we can regard country A as lending to country B to finance its exports.

Germany, China and Japan can therefore be regarded as lending directly or indirectly to the US. This lending takes several forms: purchases of US Treasuries, purchases of US businesses and real estate, cross-border lending directly into the US economy, and purchases of securities. This last form was particularly important in the years before the financial crisis: German banks had substantial investments in US mortgage-backed securities and their derivatives.

German, Japanese and - above all - Chinese lending to the US is the so-called "savings glut" that is widely blamed for creating the financial instability that led to the financial crisis. But Piketty argues that this source of capital is tiny compared to that generated by rising inequality WITHIN the US (my emphasis):
"...this internal transfer between social groups (on the order of fifteen points of USnational income) is nearly four times larger than the impressive trade deficit the United States ran in the 2000s (of the order of four points of national income). The comparison is interesting because the enormous trade deficit, which has its counterpart in Chinese, Japanese, and German trade surpluses, has often been described as one of the key contributors to the “global imbalances” that destabilized the US and global financial system in the years leading up to the crisis of 2008. That is quite possible, but it is important to be aware of the fact that the United States’ internal imbalances are four times larger than its global imbalances."
So the total amount of capital available for investment in the US at this time was far larger than the imported "savings glut" caused by its trade deficit. And because the US was importing capital, all of that capital had to be invested WITHIN the US***. As I've noted already, the corporate sector was (and is) running a structural surplus. That leaves the household sector and the government sector to absorb the capital. No wonder lenders aggressively targeted those households most in need of money. They had to put that capital somewhere****, and poor households were both the easiest to lend to (because they needed the money) and gave the best returns (because they were the highest risk).  Financial innovation enabled far more of this capital than usual to find its way to poorer households: the concentration of risk that would normally have limited individual lenders' exposure to poorer quality borrowers was dispersed across the globe through securitisation and amplified with derivatives.

And no wonder government encouraged lending to poorer households and riskier borrowers. The alternative was far higher government borrowing and lower tax revenues. But this was a short-sighted policy: when the whole system crashed, unsustainable household debt was replaced with government debt, while the recession clobbered tax revenues and raised fiscal deficits. Now, government is trying to push the debt it was forced to take on in the crisis back to the household sector again by means of fiscal austerity. And the effect of fiscal austerity when the corporate sector is in surplus and the household sector is damaged is to force down the trade deficit. No bad thing, you might argue - but reducing the trade deficit entirely by means of squashing domestic demand is beggar-my-neighbour economics. If everyone tries to reduce their trade deficit by this means, no-one can.

So we can now understand the story that Piketty's paragraph is telling. During the years before the financial crisis, as top incomes continued to rise while low-to-middle incomes stagnated, capital available for investment grew. Imports of capital from countries with trade surpluses added to the capital pile, though they were not its main source. Lenders faced with trying to generate decent returns from this capital glut offered cheap money and easy lending terms to increasingly risky borrowers, who lapped it up to fund consumer spending. Consumer spending added to corporate and high net worth returns and expanded the surpluses of exporting countries, increasing the capital pile still more, encouraging lenders to lend even more on even easier terms and creating a whole new industry dedicated to finding new ways of dumping risk on the unsuspecting. It is easy to see how this became a toxic feedback loop that eventually imploded in a disastrous crash.

And it is now easy to see why Piketty argues that US inequality contributed to the financial instability that led to Lehman. You may disagree - but it is a compelling argument.

Related reading:

Capital in the 21st century - Piketty
Picking apart Piketty - Money Illusion

* I'll use Godley & Lavoie's sectoral balance equations to explain this if you really want me to. But I warn you I usually get them wrong.

** For the purposes of this post I am including equity investment in a general "lending" category, although strictly speaking equity investment is a purchase not a loan.

*** This is of course net investment. The US could, and did, invest overseas as well. But its net overseas investment position was negative because of the large inflows of capital from trade surplus countries. Therefore we can regard all US domestically-generated capital, together with the net capital inflow from abroad, as invested entirely within the US.

**** I'm aware that I'm departing slightly from endogenous money theory here. But from a macroeconomic standpoint it actually doesn't matter. All saving must be invested: all investment becomes saving. It doesn't matter whether you express this as saving preceding investment, or investment preceding saving. It comes to the same thing in the end.


  1. Great explanation, thank you so much! Wish we could bring your systemic view to a larger public and have us all broaden our view on how we are part of these larger dynamics.

  2. "When low-to-middle incomes stagnate but top incomes continue to rise, saving naturally increases." Are you sure about this? What about the paradox of thrift? In my opinion, higher income inequality c.p. leads to a decrease in demand, output and saving. This is also a powerful explanation for secular stagnation following the debt bubble.

    1. The paradox of thrift can be described thus: "when everyone is trying to save rather than spend, no-one can". Inequality causes increased saving because those with the highest incomes save more, not because everyone does: indeed if the increase in top saving is matched by increased borrowing at lower incomes there may be no change in net saving. The paradox of thrift would only apply if saving instead of spending became fashionable at all income levels.

      What I think you may be referring to is the tendency of the poor to cut discretionary spending when incomes fall or borrowing sources dry up. This does indeed take demand out of the economy, causing stagnation.

      The debt bubble was a consequence of secular stagnation, not the other way round.

  3. Thanks a lot for your answer! Maybe I was wrong to use the term paradox of thrift. What I meant is if income inequality increases, aggregate demand c.p. falls. If firms expect lower demand, they decrease their investment. Investment drives savings and therefore savings decrease. What do you think?

    1. I don't think it is necessarily true that rising inequality causes falling demand. After all, there can be rising income inequality even when all incomes are actually rising - just some rising faster than others. Under these circumstances both spending AND saving may actually increase even though income inequality is rising. We would expect this in a strongly growing economy with no financial repression (so not China or Germany!). It might be associated with high inflation.

      Rising income inequality might be associated with falling AD if real incomes for the majority are falling faster than real incomes at the top. But the problem is falling real incomes, not rising inequality per se. Curtailing top incomes to reduce inequality wouldn't necessarily mean income rises at the lower end, which is what is really needed. In focusing too much on inequality we are in serious danger of missing the real problems.

  4. You've used about 1,200 words to "explain" what Piketty wrote in under 100 words. i don't know what the opposite of a precis is, but this certainly qualifies. I look forward to the 8,000 page explanation of Piketty's whole book!

    1. Hah. Well, Piketty did say that perhaps his book should have been longer!

  5. "So we can now understand the story that Piketty's paragraph is telling. During the years before the financial crisis, as top incomes continued to rise while low-to-middle incomes stagnated, capital available for investment grew."

    "And it is now easy to see why Piketty argues that US inequality contributed to the financial instability that led to Lehman. You may disagree - but it is a compelling argument."

    The argument is not compelling if it is based on faulty data. And Piketty is already backing away from the US data presented in the book (the one containing the quote that is analyzed here).

    The argument here seems to be that due to rising inequality of incomes, "the rich" had more money to save, and those savings were forced on unsuspecting low- and middle-class borrowers (but, at least, the government role in such lending is here acknowledged).

    Where did Piketty get his US numbers? Primarily from US tax return data. That data is suspect, and indeed wrong, for several reasons:

    1. The US tax return data Piketty uses does not include private pension savings. Those savings rose from $875 billion in 1984 to $12.4 trillion in $12.4 trillion in 2012 (a more recent report puts this at more than $20 trillion today). These savings within these plans and the income within them (IRA's, 401(k)'s, defined benefit plans, etc.) don't show up on US tax returns and therefore were not counted by Piketty in his analysis of "increased inequality". Due to the monetary restrictions on contributions to those plans, I think it can be safely concluded that they belong predominately to the under 1 percent or 0.1 percent class (the latter, particularly). While its true that many, if not most, Americans under-save for their retirements, this is a significant sum to omit from your analysis. And yet, that sum, representing largely the savings of middle-class Americans is just (if not more, due to the more conservative investment strategies) available for indirect lending back to those same middle-class folks and lower-income folks. It's easy to attribute a higher proportion of savings (and indeed, inequality) when one omits the largest source of savings the lower and middle classes have. The crisis hit just as large numbers of baby boomers were approaching retirement age.

    2. While tax-deferred income under private pension plans does not appear in US tax return data, more recently municipal bond interest does (this type of interest is tax-exempt at the federal level, but must be reported on tax returns). The change in reporting policy thus tends to lead to the erroneous conclusion that income and wealth inequality increased.

    3. Due to declining personal income tax rates in the United States and due to the introduction of tax-transparent limited liability companies (LLC's) much business investment shifted from taxable "C" corporations to flow-thru entities such as LLC's (and Subchapter S corporations and general and limited partnerships). From the mid-1990's a much greater percentage of business income has been earned through these entities and thus the income reported directly by "rich" owners on their tax returns. Prior to this, much of that income was not directly reported on individual returns because it was earned (but not distributed) within the corporate solution. Failure to account for these changes means that Piketty exaggerates the change in inequality over the past two decades or so. Simply because increased accounting income is reported on tax returns of "the rich", that does not mean it is available for "savings", much less than it represents an increase in "income" and therefore "wealth" from periods when it was reported as income of a taxable corporation owned by its "rich" shareholders>

    So, for these reasons (and more), I don't find the data or the argument "compelling".

    There is more on this in today's WSJ.

  6. Note that Alan Reynolds has been attacking Piketty on similar grounds for a very long time. Here is a link to a 2006 response by Piketty (and Saez) to one of Reynolds' earlier critiques: http://eml.berkeley.edu/~saez/answer-WSJreynolds.pdf

    1. Thanks for that link.

      Actually, Piketty's response at that link does not address the points I raised above. He does tangentially address the issue of 401(k) plans (excluding, apparently all other private pension arrangements) arguing that income from these assets is included in income when distributed. However, that misses the point. It ignores the fact that, whether these assets generate current income or not, they are a form of wealth and savings. A very significant form, especially for the "non-rich". (As is other vested rights in annuities, such as social security). It is just as much a form of savings and wealth as a stock portfolio that does not currently pay a dividend. It appears that the latter finds itself into Piketty's calculation of "wealth", but private pension savings do not. And, strangely, Piketty does not seem to view any amount distributed from private pension arrangements as "income from capital". That does not make a lot of sense to me.

      And, more directly to the point raised here by Coppola, It also doesn't make a lot of sense to me to ignore the fact that the funds these pension savings represent is made available for lending to the middle class and the poor to buy houses and other things. The very fact that they are undistributed makes them available to do just that.

  7. I can indeed help.....

    A leveraged hedge fund is essentially a fund that is hedging part of its portfolio with borrowed money, often as part of a long/short strategy. You could argue that this is what commercial banks do, actually: they borrow at a low rate, lend at a higher rate and make a profit on the spread. Though we don't usually think of banks as hedge funds. Perhaps we should.

    Banks are by definition highly leveraged - a bank that wasn't leveraged would be lending only from own equity (we don't have any banks like that at the moment, though some people would like to see them). Capital requirements attempt to limit banks' leverage, but even so the leverage of a typical bank is far higher than any corporation or indeed most leveraged hedge funds. LTCM, which was far more highly leveraged than the vast majority of hedge funds, had a leverage ratio of 4% when it failed, which is about average for a bank.

    New Century Financial was a mortgage originator specialising in subprime lending. It originated subprime mortgage loans, bundled them up and sold them on. Its business model depended on continual sales of subprime mortgage securities to free up capital and generate funds for further lending. Northern Rock in the UK had a very similar business model. This is nothing like a hedge fund, though - the securitisation engine is more like a GSE. Really, NCF and NR were both acting like private sector versions of Fannie Mae.

    Leverage is a meaningless concept for a central bank. They don't borrow. They issue currency ex nihilo and they back it with various categories of reserve assets, including their own governmment's debt, the debt of other governments (FX reserves), gold and (sometimes) private sector assets. The liabilities of a central bank are the monetary base, which behaves more like equity than debt: it is undated, zero-coupon and callable (the Fed can withdraw it at any time without notice), and you have no right of redemption - you just try redeeming a banknote and see what you get.

    Does that help?

  8. New Century Financial should be considered a "shadow bank" and not an "ordinary bank", right?

    ---Fed Up

  9. No, it was a specialist mortgage bank. Like Countrywide - which also went bust in the crisis and was acquired by Bank of America.

  10. http://www.investopedia.com/articles/07/new-century.asp

    "2004 - Converts to a real estate investment trust; listed on the NYSE; originates $42.2 billion in mortgages"

    Why is New Century Financial an "ordinary bank" and not a "shadow bank"?

    Which brings up the questions:

    What is an "ordinary bank"?

    What is a "shadow bank"?

    ---Fed Up

  11. a Century Financial was a mortgage originator specialising in subprime lending. It originated subprime mortgage loans, bundled them up and sold them on. Its business model depended on continual sales of subprime mortgage securities to free up capital and generate funds for further lending. Northern Rock in the UK had a very similar business model. This is nothing like a hedge fund, though - the securitisation engine is more like a GSE. Really, NCF and NR were both acting like private sector versions of Fannie Mae


Post a Comment

Popular posts from this blog

WASPI Campaign's legal action is morally wrong


What really happened to Signature Bank NY?