The stickiest price

Here’s a question for the macroeconomists.

“Sticky prices” are the foundation of “Great Moderation” monetary policy, the core justification for why we have inflation stabilizing central banks. As the bedtime story (or DSGE model) goes, if only prices were perfectly flexible, markets would always clear and the great equilibrium in the sky would prevail and all would be right and well in the world. Hooray!

Unfortunately there are… rigidities. Shocks happen (economists are bashful about that other s-word), and prices fail to adjust instantaneously. Disequilibrium persists or oscillates and all kinds of complex dynamics occur, because the system, once outta whack, doesn’t get back in whack very quickly. Disequilibrium is followed by its terrible twin distortion, which shrieks through the night, ravaging the villagers with suboptimal resource utilization, most especially suboptimal utilization of the villagers themselves who are let to starve because their wage expectations are too damned sticky.

If my tone betrays a certain disdain for this account, that is because, in my view, central bankers have used it to harm people and blame the victims. The policy regime that we have crowed over from Volcker through Bernanke and Trichet “naturally” led to the conclusion that (1) central banks should stabilize inflation, so that predictable price adjustments are mostly sufficient to keep things in equilibrium; and (2) that central banks ought to focus especially on stabilizing the stickiest prices, leading to distinctions between overall and “core” inflation. Among the stickiest prices, of course, is the wage rate. In practice, from the mid 1980s right up through 2008, the one thing modern central bankers absolutely positively refused to tolerate was “inflation” of wages. God forbid there be an upcreep in unit labor costs, implying that a shift in the income share away from capital and towards workers. Central banks jack up interest rates right away, because what if the change in relative prices is a mistake? We wouldn’t want that to stick, oh no no no no no. But when the capital’s share of income shifted skyward while deunionization and globalization sapped worker bargaining power? Well, we learned the meaning of an asymmetric policy response.

Even today, now that it has all come apart, economists maintain a laser-like focus on the stickiness of wages. Why can’t Greece compete? Because its “cost structure” has grown too high. In English, that means people expect to be paid too much. The solution is “adjustment”: workers’ real wages must be reduced to restore competitiveness. American economists, following in the footsteps of Milton Friedman, trumpet the glory of floating currency regimes, with which one can reduce the wages of a whole nation of workers with a single devaluation (and without the workers having much opportunity to object). The Greeks, of course, must suffer, because they are part of a fixed currency regime, and workers and employers are unable to organize the universal wage collapse that would be good for them in the way of vegetables at the dinner table.

Now, not all economists are heartless. Left economists love workers. They urge governments to devalue if possible, to chop the broccoli into chocolate cake and hope that nobody gags. These economists rail against the fixed exchange rates, because nominal wages cuts usually occur only alongside the human tragedy of unemployment. They beg governments, if they can, to just borrow money and pay workers their accustomed wages (to do some important thing or another) and hope that things work out well.

But it is always about the workers. Workers are the core problem. Macroeconomic policy, as a practical matter, is mostly about finessing “rigidities” associated with workers’ stubborn wage expectations.

Yet there is an even stickier price in the economy, a price economists have mostly ignored although it is at least as ubiquitous as wages. The price of a past expenditure, the nominal cost of escaping a debt, is fixed in stone the moment a loan is made and then endures in time, perfectly rigid, while the economy fluctuates around it. It is certainly a price, but can only be made flexible via bankruptcy — a disruptive institution, rarely modeled by macroeconomists, and rarely deployed at scale. Surely, the price of manumission must be as nimble as the price of petrol if the economy is to keep its equilibrium while being battered and buffeted by shocks.

This is an odd way of putting things, but no great insight. Everyone knows that we are loaded to the gills with debt, the real burden of which has grown as the business cycle turned. Disinflation has left us teetering on the edge of mass default and deflationary spirals, distortion, depression, destruction. The holograph sputters to life and Princess Leia implores, “Help us Obi-wan Bernanke, you’re our only hope.”

So, macroeconomists: For at least 40 years sticky wages have been a central concern, perhaps the central practical concern, of your profession. (In the models, yes, it is abstract goods prices that are sticky. But in practice, it was always and obviously about sticky wages.) You justified ending Bretton Woods gold convertibility and moving to a floating-rate regime specifically in terms of frictions associated with innumerable downward wage adjustments. Your central triumph was “beating” the inflation of the 1970s. You pretended that was a painful but technocratic exercise in monetary policy, but the durability of “price stability” had everything to do with Reagan’s breaking of union power and a free-trade regime that put pressure on the wages of all but the special. (Economists are very special, of course.) Back in the Great Moderation, central bankers chose not to emphasize the role of these political choices in explaining their “success”. It was all about targeting the interest rates cleverly, just like the DSGE models say. It was “scientific”, “independent”.

Don’t worry! I’m with you. I think unions are a poor means of supplying labor bargaining power, and wish them good riddance. I am proglobalization and free trade, or would be, if we had sense enough to subject our free trade to a balance constraint. I’ll keep your secrets. We’ll keep telling the little people that all we do is interest rates and blame whatever went wrong on Wall Street.

But here’s my question. Looking forward to the next thirty years, after we have decisively defeated wage rigidity by ensuring that the unemployed are numerous and miserable, don’t you think we should devote just a bit of our attention to tackling that other sticky price? As we reduce the bargaining power of labor, perhaps we should think about the bargaining power of creditors as well, so that if we get ourselves into a pickle where the “cost structure” of honoring debts is high, we have technocratic and politically acceptable means of managing the burden of loan contracts just as we’ve developed mechanisms to control wages.

In the 1970s and 80s, we threw away an international monetary regime and revamped the practice of central banking in order to give leaders the tools to push down hard on any upward creep in sticky wages. (Notice how there is never any talk of having Germany raise, rather than Greece reduce, its wages to “restore balance”?) Our new monetary system also made the price of escaping of some debt less sticky, specifically debt owed to international creditors foolish enough to lend in borrowers’ now-unredeemable currency. And that has helped, a lot! We’d be living in Mad Max USA already if dollar debts could be redeemed for anything other than more dollars.

But the job is not done. Domestic creditors, and international creditors who lend in their own money, still have sufficient bargaining power to make past prices stick, regardless of whether those prices remain appropriate. If renegotiating down labor contracts is hard, renegotiating down millions of debt contracts via bankruptcy is nearly impossible. Perhaps debts should be enforceable only in a pseudocurrency whose convertability to current dollars is routinely adjusted as a policy variable by the wise, technocratic central bank. Perhaps we should develop less disruptive means than bankruptcy for writing down or equitizing onerous debt. Perhaps since sticky-priced debt contracts have less rigid near substitutes called “equity”, macroprudential policy should heavily favor the latter. Put Trichet and Bernanke in a room together, and let ‘em figure it out. They’re brilliant, both of ‘em. Surely they can come up with something. But do they want to? Do they, as their models suggest, think that any pervasive sticky price is dangerous, or is it only uppity workers that trouble them?

A naive noneconomist might imagine that consistently suppressing one sticky price while assiduously supporting an even stickier price is not a way to avoid distortion, but a means of introducing it.

Isn’t it time macroeconomists stopped beating down wages and turned their attention to the stickiest price?

 
 

88 Responses to “The stickiest price”

  1. Nick Rowe writes:

    My short answer: Nope!

    1. Suppose you believed that wages are sticky, but that prices are perfectly flexible. (Some macroeconomists believe something like that, though most New Keynesians don’t). The right monetary policy is to make sure that *equilibrium* *nominal* wages don’t move, so that actual nominal wages don’t have to move. But that doesn’t stop actual *real* wages moving. If there’s an increase in labour productivity, for example, that would increase equilibrium *real* wages, the way actual real wages would increase, with nominal wages fixed, is by the price level falling. Workers should care about real wages, not nominal. And central banks should care about nominal wages (if nominal wages are what are sticky).

    2. Suppose that last year, I borrowed from you, and promised to pay you $100 this year. Sure, that $100 is “sticky”. But it’s not a price. Last year it was a price (the price we negotiated when you gave me the loan), but it’s not a price this year. It’s not a price like the price of labour or the price of cars. The price of labour or the price of cars affects how much labour or cars we decide to buy or sell at that price. And if those price are sticky and get out of whack that can create excess demand or supply of labour or cars. But that $100 I owe you is a transfer payment. It’s like a welfare payment or social security check. The decision it affected has already been made, last year. It cannot any longer create excess demand or supply, though it could have done last year, if it had been sticky last year. What it does do is transfer resources from me to you. It affects the distribution of wealth. It’s not a measure of the scarcity of goods that people are currently deciding to buy or sell.

    3. There may nevertheless be an argument for monetary policy to take into account the “stickiness” of past nominal debt obligations. But it’s a very different sort of argument for the arguments that are made about monetary policy taking into account sticky wages or prices. If you care about the distribution of wealth. Or about bankruptcy, etc., rather than about excess demand or supply. You might even argue that since aggregate demand may depend on the distribution of wealth, then it’s easier to prevent excess supply or demand due to other sticky prices if monetary policy takes debt into account.

    4. But what would be the best way for monetary policy to take the “stickiness” of nominal debt into account? Presumably, as a first approximation, by preventing any unexpected inflation or deflation. Keep the inflation rate as constant as you can, so people can predict it when they take out debts, and you don’t get any unexpected transfers of wealth from debtors to creditors or vice versa. In other words, inflation targeting looks pretty good. So you get the same answer by focussing on debt as you would if you focussed on sticky prices. Or you could argue that wages are a better measure of borrowers ability to pay. in which case monetary policy should target a constant rate of wage inflation, not price inflation. Or maybe nominal income is a better measure. If so, target nominal GDP, as Scott Sumner suggests.

  2. Yancey Ward writes:

    Steve,

    Why do we need to reinvent the wheel? The solution to that sticky price is bankruptcy. Bankruptcy is honest, and it has the added benefit of transferring productive assets to better management-those that are not bankrupt. Of course, we need government officials that don’t then transfer the liabilities onto the public balance sheet, but that is a political problem, not a central banking/treasury problem.

    Here is the way I look at all debt. It is simply a promise to work for someone else in the future. If society were homogenous, there would be no problems whatsoever- all debts would net out over the individual. Society is not homogenous- there are creditors and there aredebtors. If the balance of debt has reached the point where the debtors feel like slaves and are not willing to put out the effort to pay their debts, then bankruptcy is the honest recognition of this liberation. It teaches both creditors not to lend recklessly, and it teaches borrowers not to borrow recklessly. Anything else is going to fail. I wouldn’t trust Bernanke and Trichet to walk a dog.

  3. Nick Rowe writes:

    Continued:

    The job of monetary (and maybe fiscal) policy is not to make the sticky prices flexible. It’s to recognise that some prices can’t or won’t move easily to a new equilibrium. So instead of waiting until the sticky price eventually moves to the equilibrium, we do the opposite. We move the money supply to move the equilibrium price to where the sticky price is sitting. We move Mohamed to the mountain, we don’t try to use monetary policy to force the mountain to move.

    But, (speaking of Islamic banking) would the world be a better place with less sticky nominal debt and more flexible equity? Maybe it would. Should economists try to encourage this? Maybe we should. But the very first thing we should definitely do along these lines is to stop those government policies that are working in the exact opposite direction. Russ Roberts recently gave a very convincing argument that governments following the Too Big To Fail policy are giving a big implicit subsidy to sticky debt. Which stacks the decks in favour of debt and against equity.

  4. Nick — I’m gonna push back a bit. My short answer, yup!

    But it looks like I’m gonna go long, so I’ll respond to each of your points in separate comments.

    Re point 1: Nominal wages are inflexible, then yes the first best policy is to play with money to ensure that the equilibrium real wage matches the fixed nominal wage. But how does the monetary authority know what the equilibrium real wage is? As you say, standard New Keynesian models punt on this: they assume a more generalized price stickiness over a continuum of goods, and derive reaction functions and rules based on the math that results. (I wish I had my little Gali text with me to sound smarter!)

    But, if you think (as I do think) that wages are by far the most important price, you are outside of the standard models (the ones I’ve seen at least, which admittedly is a very small subset). The central bank has no means of knowing the equilibrium real wage, let alone to follow its fluctuations in real time. That is too precise a target for it to hit.

    But, central banks have a trick. Wages are not fully inflexible. They are only sticky downward! It is easy for businesses to give people wages, hard for them to give pay cuts. So, given limited information and ability to track the nominal real wage, a “reasonable” policy is to exert downward pressure on nominal wages when there is any reason to suspect real wages are getting too high. There is an asymmetric loss function: If the central bank errs by restraining wages, wages will quickly rise to equilibrium when the intervention is finished. But if the central bank errs by failing to restrain wages, then in order to restore equilibrium it must tolerate an unexpected change in the price level, that is non-targeted inflation. So there is a perfectly good rationale for central banks to behave as (I claim) they have in fact.

    Except… what if wages are very sticky downward and a little bit sticky upwards? Obviously no price is perfectly flexible, there are time constants and frictions associated with any change. Then central banks might still err on the side of wage restraint, but there is a hazard. If they do too much erring on the side of wage restraint, while they are not contracting, the equilibrium wage is never achieved at all! That is, if wages are asymmetrically sticky downward, but a bit sticky upward, the “right” policy is to suppress wages periodically when they fear above-equilibrium pricing, but not so frequently as to prevent rising from below-equilibrium pricing that may have resulted from their errors. But central banks can’t observe real and equilibrium wages in real time. If they restrain too frequently, they harm the economy by never letting wages achieve equilibrium. But if they restrain infrequently, they risk an inflation they either have to accommodate (violating their price stability mandate) or push against (provoking a slow-down or recession (violating their full employment mandate).

    Thus there is an asymmetric loss function for central bankers (as opposed to the asymmetric loss function in output or utility we saw before), and the optimal policy for the central banker is to restrain wages too frequently. That damages overall welfare, but it ensures they achieve both of their mandates. (I’m being chauvinistic and taking the US Fed’s dual mandate as normative.)

    Central banks do care about nominal wages, and they care to keep them set at or below the cash flows that would achieve the equilibrium real wage. So long as that is so, there will generally be full employment (or if incomes are incompatible with full employment easy credit conditions — expansion within the purview of the central bank can achieve it). And so long as nominal wage grows more slowly than productivity growth + planned inflation, 2/3s of the work of price stability is taken care of (using the 2/3 as a rough average unit labor cost).

    Of course wage growth that lags productivity + inflation leads to a loss in labor’s share of output. Which is precisely what we’ve seen since the era of quasi-New-Keynesian “Great Moderation” central banking.

    [You might claim I have contradicted myself by saying that if there is a shortage of aggregate demand due to subequilibrium real wages, the central bank can promote a credit expansion. After all, isn’t “expansionary” monetary policy exactly the opposite of “contractionary” monetary policy, so wouldn’t attempted credit expansions cause lead to wage inflation? But that is a question of timing. A monetary contraction pulls back on economic activity both by increasing the cost of capital and reducing access to business and consumer credit. A monetary expansion decreases the cost of capital and expands access to business and consumer credit. If access to consumer credit during a monetary expansion increases faster than the process of first hiring the unemployed and eventually bidding up wages in a “tight” market, then aggregate demand can rise before wages. This creates a periodicity in the behavior of wage-inflation-averse central banks: they provoke a slowdown or recession, ease, watch demand leads employment and wages, and wait until wages threaten to rise to anywhere near productivity growth. Then they contract. Labor bargaining power determines the frequency of this cycle. Central banks contract less frequency when wages rise more slowly due to deunionization, global competition or whatever, even under “full employment”. “NAIRU” falls with labor bargaining power. I think the cycle I’m describing is precisely what we’ve experienced for almost 30 years. I don’t think it was planned or intended (e.g. I don’t think central bankers knew in advance that easing would expand demand via the credit channel more quickly than it leads to wage inflation under present labor arrangements). But it was very well learned.]

  5. Alessandro writes:

    One more brilliant article. I know can grasp much more clearly the whole central banking concept (aka scam). Thanks!

  6. Nick — Re point 2.

    The cost of escaping debt is a sticky price, or at least a nonequilibrium rigidity closely analogous to a sticky price.

    In an uncertain Arrow-Debreu world with complete markets, heterogenous endowments, no restrictions on contract size, and risk-averse agents, no financial position would ever look like a bond, and no one would ever finance a purchase with a bond, except in odd special cases. It is only correct to say that escaping a debt is not a price if you posit risk-neutral agents with time consistent preferences and rational expectations. If you do, I’ll cede your point. At time zero, such agents might take bond-like positions to finance uncertain purchases, and over time some would gain and some would lose but in aggregate those would just be transfers that are foreseen realizations of an ex ante equilibrium.

    But in a semi-realistic model — maintaining idiotic conveniences like rational expectations and time consistent preferences but adding risk aversion and heterogeneity of initial endowments — agents do not hold bonds or finance purchases with conventional debt. They buy and sell insurance.

    Incomplete markets and imperfect information create a rigidities entirely analogous to sticky prices. At any point in time, under fully informed complete markets, agents would trade financial contracts to acquire state-contingent positions that offset the risk of their real assets. Straight bonds are an artifact of informational problems, transaction costs, legal uncertainty, and regulatory constraints. (Note there is no speculative disagreement under perfect information.) Since bonds are not optimal, if financial contracts are restricted to bond-like instruments, there will be distortion ex ante (as Pareto improving trades of state-contingent contracts won’t occur) and distortions at all future times (regardless of which states occur over time, who gets lucky and who gets hurt, there will be trades agents would make but cannot because they are restricted to debt-like securities).

    Just as when nominal prices are sticky, it may be welfare improving for the central bank to tilt things so that the real equilibrium is consistent with nominal prices, it may be welfare improving for some institution to “unstick” debt contracts, by turning them into the instruments agents would have chosen had markets been complete, either by directly altering the (nominal) meaning of the contracts or indirectly by changing the real value of the fixed nominal payoffs to match the state contingent outcomes agents wanted ex ante. This wouldn’t be covert. Agents would understand in advance that the real outcomes from “debt” contracts they choose would be “tilted” to match the outcomes of estimated optimal state contingent contracts (just as workers now understand that their employment, both in terms of availability and wages, is affected by central bank policy, hopefully for the best).

    Of course, neither fully informed complete markets nor an institution that simulates what would have obtained under complete markets for a billions of idiosyncratic contracts is remotely possible. Per above, I claim that central banks’ rules of thumb with regard to goods price stickiness amounts to managing wages. (I should have gone farther and specified that it amounts to managing median rather than mean wages.) That is not as good as somehow undoing price-stickiness on a product by product basis, exactly restoring the optimal equilibrium under flexible prices. But, although the “theory of the second best” reminds us that we cannot be sure, we hope and collectively claim that crude central bank management of aggregate prices is better than simply tolerating a distortion we cannot practically undo. Similarly, it is not difficult to imagine a “second best” strategy for making debt contracts more like the contingent claims agents would choose if it were practical to trade them.

    The most obvious centralized strategy would be somehow to compute an aggregate of “real debt burden”, just like we compute aggregates of inflation. Computing such a debt burden would be an exercise in arbitrary choices and very contentious, just like computing price indices. This debt burden would take into account both the inflation/deflation adjusted value of credit/debt positions and the exchange value of real assets notionally purchased with those assets. The “central adjuster” would presume agents would have chosen financial contracts that stabilize outcomes near ex ante expected values. Thus, if there has been monetary deflation (or disinflation below expectations) and a fall in real asset values, the adjuster would write down the principal of debt contracts. Alternatively if there has been monetary inflation and an increase in real asset values, the adjuster would increase the principal of debt contracts. (One might argue that behavioral “facts on the ground” — creditors are generally more sophisticated than debtors — suggests some asymmetry. But let’s put that aside.)

    If we imagine this sort of regime, it is easy to see why I am happy to call debt a “sticky price”. People would choose state contingent contracts that stabilize the net value of financing arrangements relative to ex ante expectations if it were practical. Since it is more convenient to alter the terms of the financial rather than real leg of the swap, we can think about how much we would need to adjust the principal in order to just restore ex ante expectations of value. That gives us a variable “shadow principal” (what would have been the principal if we had insurance contracts whose payoffs and premia were accounted for as adjustments to the size of our loan). We compare this variable shadow principal to the actual principal of the loan, which is harmfully fixed, sticky.

    I want to emphasize that the “central adjuster” described above was intended only as an example, to keep the analogy between our management of debt and central bank management of the price level as close as possible. That is one very blunt, but perhaps reasonable, way of resolving the problem. Other ways include overcoming the problems that prevent state-contingent financing directly (difficult due to information asymmetry and moral hazard issues, but promoting equity financing wherever people are clever enough to manage it is “first best”) or streamlining means of writing down, equitizing, or potentially writing up principal values ex post on a loan by loan basis.

    But the fundamental point holds: debt financing as an institution is an artifact of imperfect information and incomplete markets. We take a position that is long a real asset and short cash. That position has a (near zero) expected value ex ante, and agents would prefer to insure against extreme fluctuations but cannot. It is perfectly reasonable and intuitive to describe the nominal value of the principal owed as “sticky” relative to an “equilibrium” principal value which would restore the value of our position to the ex ante expected value.

  7. Nick — I gotta sleep, so points 3 and 4 together, fast. (If I’ve gotten more incoherent than usual, it’s because I’m fading.)

    Re point 3: I think I’ve made pretty clear why I think there is a strong analogy between sticky prices in New Keynesian models and the fixed principal of debt. I won’t argue that there aren’t other problems with debt that we should care about, distributional issues etc. I’m happy to be persuaded of all kinds of problems with debt. I’m not a fan of the institution, or of the degree to which we’ve made “credit” or keystone enabler of economic activity. But I don’t think my analogy is at all misplaced.

    Pe point 4: I mostly agree — the most straightforward strategy for a central bank, both with respect to sticky goods prices and with respect to sticky debt, is to hold prices and asset values stable.

    Let’s go through that. In order to preemptively mitigate the “sticky debt” problem, central banks must manage asset as well as consumer prices. They can’t be laissez-faire about asset inflation/deflation, because (remotely rational) debt always funds some asset, and the fluctuations they need to stabilize to stay near the “first best” equilibrium are portfolios containing both assets and debt. (In theory a central bank could be tolerate correlated volatility in asset prices and the real value of total debt, so that the net position stays stable. But in practice, the two legs of the swap are unlikely to stay synchronized. They’d want to stabilize asset prices, inflation (labor share of GDP), and total credit.

    That sounds like a lot, but it is really just what you’ve said. The simplest solution is to ratify ex ante expectations so that rigidities that prevent first-best response to volatility are never tested.

    But central banks are unlikely to always succeed at that, especially over the long term. If debt is a sticky price that would be state-contingent but for rigidities, then when asset prices and the burden of debt in terms of wages threaten to fall out of sync, there would be a role for central banks or some other institution to intervene.

  8. Nick — re continued:

    In theory you are right, we try to keep real variables consistent with sticky nominal, rather than shove at sticky nominal variables that are out of sync.

    In practice, I think you are wrong, per my response to Point 1 above. In practice, central banks are just not omniscient enough to adjust not-quite-observable real variables to an everchanging hypothetical equilibrium consistent with the prices we believe are sticky. They are not that smart. They simply push down on wages to fight inflation, make use of the credit channel to promote employment, and hope that wages are sufficiently flexible upward to prevent their heuristic from being too damaging.

    I agree entirely re stopping government policies that privilege debt over equity (see e.g. here).

  9. anon writes:

    There’s a world of difference between price stickiness of new contracts and price stickiness of existing contracts.

    The issue of price stickiness in general, including wages, usually falls into the first category.

    The question of price stickiness on debt usually falls into the second category.

    It’s a false comparison.

  10. Nick Rowe writes:

    Steve: I’m going to have to reply in chunks too!

    1. “But how does the monetary authority know what the equilibrium real wage is?”

    Good question. “With difficulty” is the short answer. If nominal wages were absolutely fixed it would be very hard, without a correct model of the economy. But if nominal wages do adjust towards equilibrium, but slowly, there’s a way for the monetary authority to grope its way towards the truth. If it sees nominal wages start to slowly rise, it figures that equilibrium real wages must be above the actual real wage, and so it tightens monetary policy to bring prices down. Vice versa if it sees nominal wages start to fall.

    More generally, if some price P is sticky, and adjusts slowly towards its equilibrium P*, and the Bank observes P, but does not observe P*, this is what the Bank does. It wants to keep P equal to P*. If it sees P begin to rise, it figures P* must be above P, so it tightens M, to reduce P*, and keeps on slowly tightening M until P stops rising. Conversely if it sees P fall, it figures P* must be below P, so it loosens M, to raise P*, and keeps on slowly loosening M until P stops falling.

    The big dangers are: adjusting M too slowly, so P and P* diverge more quickly than you can bring them back together; or too quickly, so you overshoot and make P and P* diverge in the other direction. It’s the lags that make it hard. If M had an immediate effect on P*, which in turn had an immediate effect on which direction P is moving, any fool could do it. Just like any fool can keep his car at a constant speed, just by watching the speedo and moving the gas pedal up or down. He doesn’t need to be an automotive engineer. Unless there’s a lag in the speedo and lag in the gas pedal. (Old metaphor).

    But there’s another way to use trial and error, if the Bank has experience. Suppose it sees P rising. It also looks at a lot of other indicators as well. The Bank asks itself: “Last time the economy looked like this, what did we do? And did we get it right, in hindsight?”. If we got it right last time, let’s do the same thing again. If we adjusted M too little, lets be a bit more aggressive this time. If we overshot, lets be a little less aggressive. With enough practice, we can figure out how to keep a car on the road, turning the steering wheel just the right amount, even if we don’t understand how it works.

    And then there’s macroeconomic models.

  11. Nick Rowe writes:

    continued:

    1b (further down your first comment on 1).

    You lost me a bit here. I’m going to take a stab at it.

    Suppose P (you can interpret P as “Nominal Wage” if you like) is very sticky downward, but partly sticky upward. So if P is above P*, P never falls. But if P is below P*, P slowly rises. That’s a problem. If the Bank sees P rising, it knows that P* must be above P. If it sees P constant, it can’t tell whether P* is equal to P or below P. So there’s potentially an asymmetry in the Bank’s response to gaps between P and P*.

    Central banks worry a bit about this possibility. Absolute downward nominal wage rigidity. That’s one of the reasons (fear of liquidity traps is another) why inflation-targeting central banks target 2% inflation, rather than 0%. If P* rises at a steady 2%, and is expected to keep on rising at 2% in future, P will eventually adjust to equal P*, and P will thereafter rise at 2% too. (At least, that’s the assumption behind the expectations-augmented Phillips Curve, and theory and experience seem to bear it out. Which means that Pdot = expected P*dot + beta(P*-P). “Pdot” means “rate of inflation of P”.)

    So if P is rising at less than 2%, the Bank figures that P* is below P, and loosens M. So the response is symmetric, at least locally. Until the financial crisis, people figured that 2% was a big enough margin for error to handle this case.

  12. BSG writes:

    Steve – re. your exchange with Nick: would you kindly address the madness of a system in which something as central as money is constantly subject to manipulation, and that has outsourced debt creation at its core to boot? Yes, I know that’s a loaded “question”. :-)

  13. Nick Rowe writes:

    Point 2. OK. I think I am following you better now. I misunderstood you the first time. Maybe because you are overstretching the metaphor of sticky prices by extending it to debt; or maybe I was just interpreting that metaphor too literally.

    Given all the informational asymmetries etc. that explain why nominal debt exists, we are in the world of the Lipsey/Lancaster Theorem of the Second Best. (That Theorem is on the Index, by the way; we are not supposed to tell non-economists about it). How should monetary policy respond to this, to try to bring the economy closer to what agents would have agreed on in a first-best world?

    Good question (if that’s what you were asking, so let’s assume it is what you are asking).

    There will be limits to what central banks can do to help here, since monetary policy can only target one variable (which could of course be a weighted average of many variables). And it must be a nominal variable (it must have $ in the units).

    Two sensible but different answers come to mind:

    1. Target nominal GDP. This makes nominal debt more like equity in the whole economy. If real GDP falls, the price level rises, and the real value of debt falls, just as equity would fall. If real GDP rises, the price level falls, the real value of debt rises, just as equity would. The lender takes on the risk of aggregate fluctuations, not the borrower. Or target a weighted average of the price level and nominal GDP, to share the risk between them.

    2. Your answer? Target real asset prices. Again, that makes debt more like equity in aggregate real assets. Should reduce bankruptcy of leveraged people and financial institutions. Makes sense.

    Now I think I understand you better, I’m going to have to re-read you and think.

    Yep. Good post, now I’ve figured out what you were trying to say, I think.

  14. Nick Rowe writes:

    OK. On re-reading, I see you are now (I think) contemplating the Bank having 2 instruments, so perhaps it can hit 2 targets. The first instrument is normal monetary policy. The second instrument is to adjust money as the unit of account for deferred payment (debts). The central bank can re-write all old debt contracts.

    Must go. I’m going for a long drive in an Aston Martin. Not mine, unfortunately!

  15. anon writes:

    So what’s the human capital equivalent of systematized debt default, which is effectively what you’re promoting?

    I think its firing people. What goes around comes around.

    BTW, the accommodation of debt stickiness is the purpose of equity.

  16. anon writes:

    “Incomplete markets and imperfect information create rigidities entirely analogous to sticky prices.”

    That’s why capital structure includes a continuum/distinction from debt to equity.

    Presumably no need for differentiating capital structure in Arrow-Debreu, but that’s not the real world, so how is it relevant?

  17. anon writes:

    “it may be welfare improving for some institution to “unstick” debt contracts, by turning them into the instruments agents would have chosen had markets been complete, either by directly altering the (nominal) meaning of the contracts or indirectly by changing the real value of the fixed nominal payoffs to match the state contingent outcomes agents wanted ex ante.”

    Again, not compatible with the fact that there is a purpose in differentiated debt/equity capital structure. You seem to want a world without risk.

  18. anon writes:

    “But the fundamental point holds: debt financing as an institution is an artifact of imperfect information and incomplete markets. We take a position that is long a real asset and short cash. That position has a (near zero) expected value ex ante, and agents would prefer to insure against extreme fluctuations but cannot. It is perfectly reasonable and intuitive to describe the nominal value of the principal owed as “sticky” relative to an “equilibrium” principal value which would restore the value of our position to the ex ante expected value.”

    The insurance for the creditor is the equity buffer. The creditor is long the asset, and long a put spread on the asset (equity).

  19. Indy writes:

    1. Whatever happened to those automatic contingent convertibles? Dead in the water?

    2. What about sticky real prices – as in the contract law remedy of “cover” (I suppose liquidated damages is the closest to ‘nominal stickiness’) You promised me 100 Widgets on the 8th, you breached, I had to go out and get 100 Widgets at the best going rate, you owe me whatever that cost. Would a “real” debt regime be any better or worse?

    In a way, secured debts are a “real debt” regime, if you don’t have much prospect of receiving any deficiency if you turn out to be under-secured (such as in non-recourse mortgages). You loaned a machine, or money to buy a machine, and upon default, you get back your machine.

    3. My favorite sticky price example, check out that price history: http://www.redfin.com/WA/Seattle/7500-S-Taft-St-98178/home/177690

  20. anon writes:

    TIPs are real debt.

    Secured debts are not real debt. They are idiosyncratic (price) hedges.

  21. David Pearson writes:

    Steve,

    Debt is the legacy of past asset prices. As long as it has a “sticky” price, it depresses future real returns on assets. That is, an overvalued mortgage on a bank’s books represents a house that is still being valued at the legacy price. To achieve a new, “clearing” price, the mortgage would have to be written down so that the house can be sold at that new price.

    Debt values take time to adjust–they are near-term “sticky”. In the presence of bail outs, however, that stickiness becomes long term. The result is that long-term return projections for investment projects are both low and uncertain (as they depend on bail-out policy). This depresses investment, reduces potential supply, and makes the economy more fertile ground for producing inflation.

    Economists in general seem to assume that the economy presents a consistent set of returns to investment that result in “trend growth” (because they think technology cannot be “forgotten”, I suppose). They seem to discount the concept that legacy prices (in the form of debt value) determine future expected returns, and that future expected returns are what determine trend growth. My fear is that they will attempt to pistol-whip the economy back to that trend line without allowing debt prices to adjust, and that this will result in all manner of future economic ills.

  22. Jim Bradley writes:

    What amazes me, is why anyone thinks this (attack on workers) isn’t baked into the cake when we allow a special protected class of entities (banks and the Federal Reserve) to PRINT (figuratively) money while everyone else has to work for it. Really, every loan extended is socializing the costs while privatizing the gains … the value extended comes from dilution of purchasing power from existing money holders and the interest spread earned goes to the bank at the expense of savers that would have higher yields. And as a side effect, so much for gently falling prices and real incomes rising…

  23. Andy Harless writes:

    David,

    I understand how bailouts (given that they are not certain to happen in the future — although I should point out that, even if there are no bailouts in the present, future bailouts remain a possibility, so its not clear that avoiding bailouts reduces uncertainty) make long-term return projections uncertain, but how do they make those projections low? If anything, I would expect the opposite: greater uncertainty makes investors less willing to invest at any given expected rate of return, so fewer projects are undertaken, and the marginal product of capital rises.

  24. Andy Harless writes:

    Steve,

    I think Nick may have already said this and said it more rigorously than I will, but I don’t understand your argument that clamping down on rising nominal wages will result, over time, in lower real wages. If nominal wages are rising and the central bank tightens, the immediate result will be downward pressure on goods prices, not on wages, since the goods market is more directly affected by monetary policy than is the labor market. The result would most likely be an increase in real wages, at least in the short run. For a small, open economy, this is even more dramatically true, as central bank tightening results immediately in a reduction in the domestic-currency price of foreign goods and hence an increase in domestic real wages. Somehow, you seem to be arguing, by raising real wages in the short run, the central bank run is lowering them in the long run.

    And if real wages are so low after years of being systematically reduced through this paradoxical process, how is it that businesses are not taking advantage of this cheap labor by hiring aggressively? If anything, the current situation suggests that median real wages are much too high, whereas average real wages may be only a little bit too high. Businesses complain about the lack of availability of high-quality skilled labor (people on the upper tail of the skewed wage distribution) while refusing to hire any of the millions of median-wage earners who are desperately seeking employment. (Admittedly, there is reason to be suspicious of anecdotal evidence, but my sense is that there really is a relative — though not absolute — shortage of high-quality labor.) Rather than resorting to paradoxical theories about central bank policy, isn’t it easier to explain the stagnancy of median real wages by pointing to a relative increase in the demand for more skilled labor — an increase which would be consistent with observations about advancing technology and increasing competition from low-skilled labor in developing nations.

  25. Yancey — In theory I’d agree with you, but in practice bankruptcy just hasn’t worked in dealing with the macroeconomic as opposed to microeconomic issue. Debt is still to rigid. We consider it sufficiently reasonable for creditors to expect set fixed payments that we can’t help but bail them out. And despite both poor lending and macroeconomic upheavals that might upset even relatively prudent debtors’ plans, the terms of bankruptcy are sufficiently disruptive and difficult that debtors often remain harmfully on the hook rather than going to the courthouse to restructure. If the “masses” of debtors for whom current conditions are terribly unexpected and difficult did go to the courthouse, we could afford to handle the load.

    Fixed credit has its advantages (it’s simple, requires relatively little work on the part of creditors), but over all I think the harm of that rigid price is very high (relative to arrangements that would involve ex post risk sharing). We could reform/streamline/living-will/prepack bankruptcy to make the process macroeconomically relevant and less disruptive, and I’m fine with trying that approach. But one way or another (I claim), we have to make the process of equitizing credit easier.

    We are losing years of economic activity because so many households and businesses are caught at the precipice of bankruptcy, neither getting it over with, nor making full use of their economic capacity for fear of taking risks that would push them over the edge. That is just dumb.

    Also “bankruptcy” sounds better defined than it is. Business bankruptcy is very different than personal bankruptcy, as are the different “chapters” of bankruptcy. Savvier individuals put businesses between themselves and their liabilities, but it is not a particularly just outcome that heavily indebted ordinary households face more painful “resolutions” than people who are clever about compartmentalizing liability.

    This was more than I wanted to write. To sum up, if bankruptcy had worked, if we hadn’t bailed out everybody and if insolvent firms and households had been quickly resolved and we moved on, I’d totally agree with you. But it hasn’t worked, so we either need to streamline it or try something else.

  26. Alessandro — Thanks! More of a cartoonish rant, but I’ll take all the compliments I can get.

  27. BSG — Alas, for something as central as many, we’d like there to be some kind of fixed point, an immutable, reliable unit of value. But no such thing exists, or can exist. Human wealth is ephemeral and contingent.

    I’m definitely a malcontent with respect to our current monetary arrangements. I think central bankers are nice technocrats who mean well but who have played a role in some really harmful social and economic developments, as well as accommodated a lot of corrupt profiteering from banks and the “rentier class” generally.

    I think we want our economic system, including our monetary arrangements, to be based much more on equity than “credit”.

    Are those loaded enough answers?

  28. anon — “There’s a world of difference between price stickiness of new contracts and price stickiness of existing contracts.” You assert that, but don’t justify it. I think you are wrong.

    There’s a technical sense that Nick brought up in which you are right. You can construct certain kind of models in which debt contarcts would be equilibrium results including their fixed prices. “Stickiness”, in the sense it is used in developing “New Keynesian” models, implies something that keeps a system out of equilibrium. So to the degree that your model or economy generates debt contracts as an equilibrium outcome, “there is a world of difference” as you say.

    But, as I argue in the conversation with Nick, conventional debt contracts are not an equilibrium outcome of models that are rich enough to remotely resemble reality and that fulfill the conditions under which we’d know an equilibrium to be “Pareto optimal”. You can eliminate either of those conditions to get debt contracts: you can come up with overly schematic models or models with risk neutral agents and get straight debt. Or you can presume imperfect information and get speculative disagreements and end up with straight debt. But in the former case, your model is to stylized to be useful, and in the latter case, all you show is that debt sometimes happens, not that it’s optimal or useful.

    But all of this is too technical. If we relax our assumptions realistically, and leave the world of stylized New Keynesian models entirely, debt very clearly is a price. We consider homes and rents to have prices, even though all of us are born short housing. We did not optimally and voluntarily choose our short position in housing, but do you want to argue that most debt is optimally and voluntarily chosen by rational optimizers? Once in place, escape from debt is a highly sought good, and it has a price.

    If we (realistically) consider humans without time consistent preferences, who value the future more than the past and the present more than the future, then the fact that “yesterday’s you” went on vacation and bequeathed you a big debt implies no overarching optimality to the fixed price. You just get stuck in a world where you have to pay or suffer consequences (you do care about the future, remember) and that payment involves a price. That price is rigid, and it could well be, and sometimes obviously is the case, that the price is harmful in any sense you choose: harmful to you, Pareto destructive, aggregate welfare destroying by the usual ways you might fake an aggregate welfare measure. Whatever. Note that both creditors and debtors can be harmed if the price is too high, yet it may not be trivial for parties to renegotiate given real-world complexity (precedents with respect to other borrowers, the creditors own indebtedness and need to maintain the balance sheet value of assets, etc). It is entirely possible that all parties would be better off had they entered into arrangements where the burden of the debt varied with the circumstances of repayment. It’s not just a zero-sum tug of war between creditors and debtors.

  29. Nick — I’m sensing that we might agree more than we disagree. Which is a problem, for you. I’m a total crank about central banks. As we speak, I am in an undisclosed location by the Black Sea hiding from the Fed’s Illuminati assassins. (You do know why they that big pyramid eye thing on US money, don’t you? Aluminum foil is the only defense!)

    Anyway, we are pretty close to agreeing about what central banks do in practice. They watch prices, draw rule of thumb inferences, have theories about equilibrium and hope for the best. Mostly they learn by trial and error, which necessarily means that their behavior depends upon how they evaluate success and failure. But their criteria of success and failure can’t be based on accurately targeting a hypothetical equilibrium. That’s unobservable. But their criteria are well known: stable prices and full employment. Do you not think trial and error based on these evaluated criteria might create incentives to suppress real wages and expand credit? After all, most of the price of goods and services is labor, and if markets are competitive, even in the face of ample credit, prices will stay pretty stable if labor costs stay stable. If credit expansion can produce full employment despite suppressed wage income, why wouldn’t that look like success to a central banker? Equilibrium is a bull’s eye, hard to hit, while below equilibrium wages comprise an easy to hit interval. Central bank fine tuning of credit expansion is much easier — very directly affected by monetary policy and likely with shorter lags. My core claim is that trial and error (helped along by the enthusiasm of creditors and financiers) led central banks to an apparently good strategy for achieving their mandates. Do you think this didn’t happen?

    I appreciate, and read carefully, your detailed account of central bank heuristics for achieving equilibrium (rather than for taking the short cut I claim they have taken). I think those are reasonable heuristics, and probably capture a lot of what central bankers think they are doing. (I hope they don’t imagine themselves to be overtechnically optimizing some DSGE model.) But there’s a lot that can go wrong in those heuristics. We don’t really know if P* exists, our measures of changes in P are noisy, P itself might be noisy, if P* exists, does targeting it change it? Of course the whole point of the DSGE style of modeling is to persuade ourselves that there are actions we can take that won’t corrupt our relationships, the whole edifice is a reaction to the Lucas critique. So we have our “microfounded” model of optimizing agents, and we derive heuristics similar to those you describe as central bank reaction functions which provably, demonstrably, lead to the stability and output characteristics that we’re after if applied. So we don’t have to worry any more whether pushing up the real price level of the nominal price level is trying to slowly drifting up will itself scatter P*. We have a rule that we can apply and nothing breaks. But that rule (which ends up just ratifying the heuristic you developed in 10 minutes with much greater clarity than all that DSGE stuff) is only “demonstrably” stable if we believe our model. Do you think those fancy linearized New Keynesian models are reasonable approximations of reality? Note we can’t rely on Milton Friedman here and say “it doesn’t matter as long as they work”. Because we’ve seen that there are other strategies that will seem to work, strategies that have nothing to do with an optimal equilibrium and that arguably favor some constituencies over others and that perhaps are outright welfare destructive. So everything depends on whether you believe the model “holds” a priori.

    There are a variety of New Keynesian models that end up basically ratifying Taylor-rule-ish heuristics. Perhaps you take some solace in the fact that these results are “robust”. But then all of the models are fundamentally similar, based on the same framework and approximations that some (e.g. Willem Buiter) would argue are cartoonishly simple. It takes a lot of human judgment to decide whether multiple models supporting the same conclusion offer real information or suggest confirmation bias.

    For me, obviously, I don’t trust this stuff. The space of potential models is vast, and the basic DSGE models are not persuasive to me in their setup, and so aren’t persuasive in their conclusions. So I don’t think the Lucas critique is resolved, I think we are just going by trial and error and past relationships and have little idea where equilibria might be, whether they exist, whether they are multiple and if so how to rank them. I think the whole edifice is a thin veneer over trial and error. And you already know how I think that trial and error works out, given central bankers’ mandate and incentives.

  30. De K writes:

    I’m having difficulty understanding the notion of disequilibrium in wage rates. In the case where the central bank has pinned wages to a suboptimal level:

    Is it the case that for that price level
    – workers are supplying less labor than the equilibrium point
    – while labor is demanding more than the equilibrium point

    And therefore there is a loss to the economy equal to the gap of labor not supplied multiplied by productivity?

    It would seem that the suboptimal wage level creates an unnaturally high “natural rate” of unemployment, then.

    My macro is dusty…. And this conversation is interesting enough to risk rousing my allergy to frank expressions of ignorance.

  31. VJK writes:

    Steve:

    I think we want our economic system, including our monetary arrangements, to be based much more on equity than “credit”./i>

    In fact, the bulk of credit is *already* issued in the form of equities and bonds. Such credit no different from endogenous bank credit creation, the only differences are that there fewer constraints wrt. liquidity and capital, and no government support (FDIC, etc).

  32. […] Tabarrok)• The capital tsunami is a bigger threat than the nuclear option (Michael Pettis)• The stickiest price (Steve Waldman)• Living with a computer (The Atlantic, from […]

  33. Nick — You economists, you keep your “index” pretty well hidden. I’ve taken all the coursework for an econ masters and a finance phd, and never encountered Lipsey/Lancaster until a blogger, Yves Smith, turned me on to it. I tried to allude a bit to that paper in one of my earlier comments, but was a bit oblique. But yes, I think that’s very much where we are. We are hoping for the best without the math to prove it. (Which is really all we ever do however much math we think we have.)

    I think you mostly get my point. The first best, cross-selling of fine-grained state contingent contracts to finance whatever it is we are up to is unavailable, impractical. (Unless, of course, we imagine an economy that includes personal relationships, which can be thought of as means of enabling desirable contracts that could not be intelligibly specified as formal legal documents!) So people enter into debt contracts. As reality evolves, the principal of the debt fluctuates relative to the first-best burden that would have occurred under complete markets without information problems or transactions costs. I see the first best burden as a “shadow price”, the fixed principal as a sticky price, and argue that is we could bring Mohammed to the mountain or the mountain to Mohammed, we could simulate the first best. Lipsey/Lanacaster remind us to be careful, that approximating first best is not the same as achieving it and is sometimes paradoxically worse. I do as the economists do, shrug and move on. Fundamentally, I think when the “shadow price” as I’ve defined it and the fixed principal are extremely out of whack, the principal should be adjusted. The difference between the principal and the shadow price depends upon the value of real assets (the stuff financed), some inflation measure, and outstanding credit. (Alternatively, it depends on difference between the real asset and the real burden of the principal.)

    I think that targeting nominal GDP is also an interesting idea. It would certainly reduce the degree to which the burden of debt relative to the nominal value of real assets changes, thereby reducing the divergence of the fixed principal and the shadow price, thereby making the rigidity relevant and destructive. But there would still be plenty of scope for asset price moves to screw debt contracts. (If the US Fed were targeting NGDP, I suspect housing prices would still have fallen and mortgage contracts would still be underwater.) So targeting NGDP might or might not be sufficient.

    Hopefully the long drive was a delight!

  34. anon — Arrow-Debreu here is just a way of talking about things in formal equilibrium terms. In an Arrow-Debreu world, all conceivable contracts are available, which is what you are getting at.

    In the real world, many contracts can’t be written (the degree of contingency would turn them into phone books), and a few kinds of contract are “standard”, such that they have a huge transaction cost advantage over the rest. Further, many contracts that would be desirable in a world without information asymmetries are spurned in a world where our counterparty might know something and wish to screw us. But, Arrow and Debreu tell us, we can only be sure of the optimality of trading equilibria if there is not information asymmetry.

    So, despite the fact that we do have a menu of contracts in the real world ranging from pure debt to common equity to informal collaboration, we won’t choose the contracts that would lead to an “optimal” equilibrium outside of an Arrow-Debreu world. That’s the problem. If we want to move towards to “Pareto optimality” in reality, we need some way of getting around the fact that real-world messiness will prevent us from making the optimal choices. In the typical sticky price story, the central bank tries to ensure that the prices we are restricted to choosing stay reasonably close to optimal. I’m suggesting we devise means of adjusting debt contracts to resemble the debt/equity hybrids we might choose in an idealized world. If a central authority universally adjusts contracts due to systematic changes in debt burden or asset prices, that eliminates the moral hazard that my own individual counterparty will lie or fail to exploit her assets in order to have her debt to me reduced. Only systemic events change the principal owed. If we want a more fine-grained solution, there’d have to be some third party (the role a bankruptcy court currently serves) to monitor the validity of revaluations.

    I’m not denying that there are differences between debt and equity, that there is a menu of choices. My claim, however, is that much of the popularity of debt is due to the difficulty of monitoring asset values and lack of trust in counterparties if they cannot be compelled to pay. If we could design equity-like contracts that overcame these problems and make them simple and standard so they could be widely used, we’d have solved the problem. Most debt financing would simply migrate to those contracts, as they would be superior financing vehicles. (I’m interested in developing such things.) But until we have solved that problem, we have to deal with what I think are terribly destructive deficiencies of conventional, straight debt.

  35. anon — “The insurance for the creditor is the equity buffer. The creditor is long the asset, and long a put spread on the asset (equity).”

    we have to consider the kind of assets financed and the sort of recourse.

    but you are right, debt often includes various forms of insurance for the creditor, it is insurance for the debtor that is difficult to come by. (we can characterize debt via options positions in a variety of ways. i’ve seen fun arguments squabbling over positions that are economically identical if you understand put-call parity.)

    conventional debt exists largely to overcome creditors’ fear of being screwed by the person who quickly takes possession of the money. creditors are given a claim that does not depend much on informationally opaque aspects of the enterprise financed, and is given a giant stick in the ability to claim collateral or force bankruptcy on nonpayment. in a “perfect information” world, creditors would be willing to accept more contingent payoffs (since they can verify the contingencies) and wouldn’t need a third party to arbitrate or force payoffs in bad states of the world (that’s just another contingency).

    in other words, in a perfect information world, creditors would be willing to offer insurance or risk-sharing to debtors that in the real world they do not offer except for a hefty return premium.

  36. Indy — Some banks are issuing contingent convertibles, and they remain trendy among banks and regulators.

    Despite having proposed such things a while back, I’m now pessimistic on the concept. (Coincidentally I explained why in an e-mail to Nick Rowe not so long ago.) But the idea is very a propos here. I just don’t think we’ve overcome the problems that make equity-like instruments difficult to issue: there is a lot of scope for gaming (by issuers or investors) or politics (when regulators are involved in the conversion decision) to skew the outcomes.

    A “real” debt regime would not be better, but thinking about it helps to see why outstanding principal is a price. We are all born with real debt, e.g. we are short food and housing. But we still acknowledge that what we are liable for has a price, and understand in an economic sense that some prices (of food and housing) are more reasonable than others.

    Why wouldn’t “real” debt (payment in kind, or cash settled based on commodity value) be better? Because the value of the financed asset portfolio can still fluctuate against the burden of the liability, whether the liability is measured in money or widgets.

  37. All — I’ve been going through comments sequentially, but I’ve got to sleep now, it’s late where I am. More tomorrow. G’night!

  38. Lord writes:

    I certainly don’t believe if prices were completely flexible, equilibrium would prevail, rather the result would be instability due to everyone trying to out anticipate everyone else’s expectations. Stickiness is friction but without it there would be no method of control at all.

    The principal is fixed but debt has become more flexible over time such as adjustable rates. Alas, in many cases, like credit cards, it has become adjustable in the wrong direction at the wrong time.

  39. anon writes:

    “in other words, in a perfect information world, creditors would be willing to offer insurance or risk-sharing to debtors that in the real world they do not offer except for a hefty return premium”

    I know you’ve effectively restated this point a number of times in different ways, but I’m still having problems in understanding what you’re getting at.

    You seem to be describing a more granular, contingent payoff structure for debt capital structure.

    But you’re still describing debt, aren’t you?

    It’s not as if you’re suggesting some complex granular multi-CDO insurance structure.

    It’s debt.

    And isn’t there a problem in ensuring that complex contingent payoff structures do not result in inconsistent and overlapping claims across the entire capital structure? Isn’t there a limit to complexity and a benefit to simplicity, even in theory?

  40. anon writes:

    I guess I was thinking about synthetic CDOs via CDS insurance above

  41. Nick Rowe writes:

    Steve: First I want to just clear up the confusion over “debt is a sticky price”.

    Steve said: “anon — “There’s a world of difference between price stickiness of new contracts and price stickiness of existing contracts.” You assert that, but don’t justify it. I think you are wrong.”

    Suppose I buy a car this year at a price P, and get a one-year loan at interest rate i to pay for it, so I promise to pay P(1+i) next year.

    This year, if P is stuck at the wrong level, or if i is stuck at the wrong level, so they don’t change to changing demand and supply, we get problems. There will be excess supply or demand for cars (or loans). So the quantity of cars sold (or loans sold) will be forced away from the flexible-price equilibrium. That’s a true “sticky price” in the normal macroeconomic sense.

    Next year, I have a debt of P(1+i), and neither P nor i can now change if economic circumstances change. You might say that P(1+i) is a “sticky price”, but that metaphor can be very misleading. Because I have already bought the car, and the seller has already sold it. So the ex post “stickiness” of P(1+i) this year cannot cause excess demand or supply in last year’s car market. So debt is not a sticky price in the normal macroeconomic sense.

    So if Steve means that sticky debt is a problem in exactly the same way as sticky prices are a problem, I disagree.

    What else might Steve mean? (What *should* Steve mean?) Either:

    1. The “stickiness” of nominal debt doesn’t create the same problems as sticky prices normally do in macro. But maybe it creates other problems, for other reasons. Bankruptcy, failure of financial institutions, etc., which create externalities and so aren’t socially efficient.
    Or
    2. We are “stuck” in an equilibrium where we borrow in the form of fixed nominal debt, rather than equity, or other types of contingent debt. Because of asymmetric info, etc. And this creates externalities which are not socially efficient. Lipsey/Lancaster.
    Or both 1 and 2.

  42. anon writes:

    Somebody needs to develop a theory of central bank inflation targeting that corresponds to the “trilemma” thing for international finance. Can central banks target inflation on economic output as well as the values of “old” assets (financial or real)? I doubt it. It amounts to targeting discount rates for present values and the values themselves, simultaneously. Must be a conflict there. Controlling income inflation amounts to the first type, since it correlates with output price inflation. Targeting the “stickiest” price, ex post, amounts to the second type. Must be a conflict there.

  43. Nick Rowe writes:

    Assume that Steve means both 1 and 2 in my comment above. We are “stuck” in an equilibrium in which a lot of loans are “sticky” in nominal terms.

    Then the rest of what Steve says makes a lot of sense.

    Even if wages and prices were perfectly flexible, for example, it would mean it would be a vary bad idea for the central bank to cause large unpredictable fluctuations in the price level. This would make the market for loans much less efficient both ex ante (neither lenders of borrowers would be happy with this exogenous increase in the uncertainty over the future real value of their asset or debt) and ex post (lots of extra bankruptcies, which can have real costs, for example).

    There will presumably be some optimal monetary policy from this perspective, that would to some extent “compensate for” the nominal stickiness in debts. Maybe (or maybe not) it would lean towards trying to re-create in real terms, at least at the aggregate level, what debt contracts would look like in a first best world. Just as optimal monetary policy in a world of (say) sticky nominal wages might try to re-create in real terms, at least at the aggregate level, what wages would look like in a first best world.

    So, given that Steve would be right on this point, then his wider point makes sense too, at least in principle. He is saying (or should be interpreted as saying) that macroeconomists pay too much attention to getting monetary policy to compensate for sticky wages (or prices), and not enough attention to getting monetary policy to compensate for sticky nominal debt.

    In practice though, I don’t think there is a lot of conflict between these two objectives. A monetary policy which solves one problem well will also be a fairly good solution to the other problem as well. Inflation targeting or NGDP targeting, would be pretty good on both grounds.

    But Steve’s point would I think give extra reasons for wanting *level path* targeting, rather than “growth rate” targeting. For example, Steve should argue on these grounds that price level path targeting would be better than inflation targeting. The reason is that if the Bank targets (say) 2% inflation, then the price level (and the real value of old debts) will be a random walk with drift. And a random walk has a variance that grows to infinity the further we look into the future. That’s a big problem for long term nominal debts.

    But, people have made this case before for price level path targeting vs inflation targeting.

    But^2, there might be other cases where sticky prices motivate a different monetary policy from sticky nominal debt.

    So, Steve, if we ignored sticky prices and wages, and just concentrated on sticky debt, what would optimal monetary policy look like? Then we can see if it looks any different from optimal monetary policy to deal with sticky prices (or wages). Then if it does, we look at compromises between the two objectives.

  44. Nick Rowe writes:

    anon: “Somebody needs to develop a theory of central bank inflation targeting that corresponds to the “trilemma” thing for international finance.”

    It’s been done. Here it is in a nutshell:

    1. If you have one instrument, you can only have one target.

    2. Given long run monetary neutrality, the thing you target must have $ in the units. You cannot use monetary policy to permanently target unemployment, for example.

    (On both these grounds, the Fed’s official objective, unless you interpret it very charitably, is incoherent.)

  45. Nick Rowe writes:

    De K writes:

    “I’m having difficulty understanding the notion of disequilibrium in wage rates. In the case where the central bank has pinned wages to a suboptimal level:…”

    If the labour market were the textbook perfectly competitive supply and demand model, and if the Fed permanently pinned the *real* wage below the equilibrium, then yes there would be permanent excess demand for labour, and employment and output would also be permanently lower. And this would be bad.

    But the Fed *can’t* permanently pin *real* wages (or any real price) below the equilibrium. If it attempted to do so, the result would be explosively accelerating inflation.

    The Fed could permanently pin nominal wages. But then real wages would still adjust to equilibrium.

    We (macroeconomists) are not talking about permanently pinning any real (relative) price. We are talking about preventing short-run fluctuations of real prices relative to their equilibrium values.

  46. rjw writes:

    slightly off topic but ….

    welcome though it is (and it is !) to see this style of macroeconomic reasoning experiencing serious attention , I’ve always had one nagging thought … how does one explain hyperinflation episodes? ( in order words, is there anything at all in the conventionl view of hyperinflation as a result of ‘money printing’).

    Intuitively, it seems to me that the onset of hyperinflation need not be itself driven by money … but that at a certain point a collapse in confidence in the currency as a store of value means that a type of quantity theory may come into play, where people try to shift out of liquid assets as quickly as they can … leading to a spiralling demand for real assets that cannot be met.

    In short, is there any MMT literature on the dynamics of hyperinflation?

  47. David pearson writes:

    Andy,

    Bail outs lower future returns by keeping asset prices higher than they would otherwise be. Returns are a function of cash flows, acquisition price and the discount rate. New home buyers face lower returns than otherwise because bailed out banks do not have to sell assets and so recognize losses. The Fed tries to make up for lower return expectations by artificially holding down the discount rate in the npv calculation.

    Japan is the poster child of a decrepit banking system depressing growth by refusing to write down asset values.

    In short debt matters because unless it is written down, the economy will suffer from low return expectations and chronically low investment in non export industries.

  48. Andy Harless writes:

    David,

    Why is it a problem that home buyers face low return expectations? Is it really so important to encourage home buying? It seems to me that low expected returns for home buyers are a good thing, because they encourage investors to find more productive channels for their investment. The bailouts were mostly housing related, but the problem is not lack of housing investment but lack of investment elsewhere. How can you blame the bailouts for that?

    And in what sense is the low discount rate “artificial”? The Fed is trying to balance the supply and demand for money. Since the marginal cost of producing money is zero, the Fed attempts to produce enough to satisfy demand, thereby maximizing the consumer surplus. (OK, it’s more complicated when there’s inflation, but when we’re already in a liquidity trap, the old Milton Friedman argument about the optimal quantity of money applies: an optimal producer of a zero-cost good will keep producing until the price — in this case, the interest rate — is zero.)

  49. David pearson writes:

    Andy,

    I am making a more general point that debt represents the legacy value of assets. If in it’s efforts to fight deflation, the fed induces banks to avoid writing down debt, then asset values are sticky and this prevents the economy from relizing “full investment”. Why believe that sticky goods prices are a problem but not sticky asset prices?

    On the discount rate, I meant the l.t. Rate used to discount cash flows from an investment. Here the fed keeps it artificially low by buying long term bonds and by promising low downside risk–again, due to bailouts. As long as these measures delay the adjustment in asset prices, they depress investment. Eventually, as was the case in Latin countries, the low savings and investment may reduce potential output and increase the tendency towards inflation.

    Over the past ten years, what portion of the time has the U.S. Economy been under some form of attempted stimulus as defined by the real rate? And yet economists continue to argue these are temporary measures and that organic “trend growth” is what we should expect. Over time, this fiction becomes harder and harder for investors–faced with actual investment projects–to accept.

  50. OGT writes:

    Nick- I think I just lost whatever remaining respect I had for macroeconomics reading your comment #44 #41.

    Let’s say we replace ‘car’ in your example with something more esoteric like say houses. A house is purchased with a loan price equal to P plus i, ya-da ya-da. The price is based on the implied rental value (We’ll call that Owner’s Equivalent Rent, where’ve I heard that before?) plus reversion value. So a shock happens, the asset value return of implied rents plus reversion value is far below the value of the debt (inconceivable!), but the debt service is fixed as is the outstanding loan amount. If all similarly placed debtors attempt to sell at the same time asset prices fall and both the debt service and outstanding loan are even more out of whack. So how is this not a fixed price?

    Or change it to Factory plus equipment. A factory plus equipment is purchased with a loan price equal to P plus i, ya-da ya-da. The price is based on the implied income stream value plus reversion value. So a shock happens, the asset value return of implied revenue plus reversion value is far below the value of the debt. If all similarly placed debtors attempt to sell at the same time asset prices fall and both the debt service and outstanding loan are even more out of whack. So how is debt and debt service not a fixed price?

    But, of course, economists have an answer… more debt, er credit, which sounds much more respectable. No, the economist might say we just print money and give it to banks what the banks do with it is their perogative. Plus not only do we have a plan, we have added a GOAL. Though as anyone who’s ever dieted can tell you having a goal and acheiving it are different things.

    This ratcheting up of debt works as AD is demand plus change in debt. Change in debt can never be allowed to be negative, or not anymore negative than needed to supress wage growth. This works until it doesn’t. It stops working when the credit channel stops working (IE there is not sufficient demand for loans). If any one doubts the debt ratchet effect, they should see the Alan Taylor paper linked below.

    No, No! The economist might say, “We work on Expectations! Expectations of future credit, er money growth.” But can people stay solvent long enough for the Fed to act irrationally?

    Put another way, as is well known in economic circles (discovered only two hundred years after the discipline’s founding), in the long run we’re all dead. Coincidentally this is the same time frame in which money becomes ‘neutral.’ In any case, how much should economic entities care about an ‘ironclad’ Bernanke promise to increase the money supply by a power of five in 3010? Um, not much. 2110? Maybe, possibly a bit, but not much, the future commitment is rather contigent and I will be long dead by then. 2040? Hmmm, a bit more relevant, but I have many more intervening interests that are more important drivers or my behavior.

    So what is the relevant time frame for ‘expectations’? Does it depend on one’s age? Well, at least one’s expected life span, if the doctor says, “You’re probably not going to live through the night, but if you do we’re serving pancakes in the morning.” How much are you looking forward to the flapjacks?

    Does it depend on your fiscal position? The Greeks and Vegas homeowners would certainly say yes. As would Japanese, and now, American business people.

    In a Balance-Sheet recession conventional means of managing the money supply via interest rates and ‘expectations’ of future credit growth are not effective. And the reason is ‘sticky’ debt and debt service.

    End Rant.

    http://www.econ.ucdavis.edu/faculty/amtaylor/papers/w15512.pdf


    [Note: After dropping some spam, a reference to #44 in the comment above had to be changed to #41 to maintain the original meaning. I’ve corrected that above. The comment is otherwise unmodified. — SRW]

  51. Inflate away writes:

    SW: Excellent post…I now officially add you to the list of DeLong, Krugman, Sumner and others who are yelling out for “higher inflation”. So that “sticky” debt payment amount becomes little more slippery and bearable. Is BB listening?

  52. BSG writes:

    Steve – Thanks! Though somewhat oblique, your answers are loaded indeed.

    It seems to me that much of the technical discussion generated and supported by central bankers is meant more to obscure than illuminate. The sordid history of banking crises, their aftermath and the creation of central banks in such a manner so as to protect and peridically save a rapacious and predatory industry seems to be all but off limits in policy discussions. I don’t think it’s a coincidence. I don’t think it’s helpful.

    I admire and have learned a lot from the analytical framework you use to attack the problem. I think it’s important to use the minions’ tools against them. I just don’t think it’s enough.

    Thanks again and keep up the good fight!

  53. VJK writes:

    David @ 49:

    Here the fed keeps it artificially low by buying long term bonds

    Fed do not control LT interest rate– for to do so they have to be both the marginal seller and the marginal buyer at the same time in the LT market. They are in the ST corner of the yield curve, namely interbank market, but in the capital and money markets there are other competing assets, like corporate bonds and commercial paper and no one is obligated to buy only government securities o Fed cannot fix price in those markets.

  54. David P @ 21 — I mostly agree with you in practice.

    But I’ll play the devil’s (Fed/Treasury’s) advocate a bit: debt is always half of a portfolio. There is the debt assumed, and the asset purchased. (Sometimes the asset purchased is an ice cream cone already eaten. But we’ll make the devil’s life a bit easier and assume debts finance reasonably matched financial assets.)

    To “unstick” the price of debt really means to resolve a mismatch between the asset and liability side of the swap. If assets values have depreciated, you can do that two ways: 1) you can write down the debt; 2) you can somehow cause the asset values to reflate.

    Bailouts are often intended to do the latter. They are, therefore, one means of unsticking the sticky price of past debt.

    I think they are a very poor means, relative to writing down the debt, and I’m guessing you agree. Bailouts tend to unstick some peoples’ debt and not others: the government picks winners and losers. Bailouts destroy incentives to choose assets well and maintain or improve their value — why worry, as long as you select assets that will be bailed out? Finally, if you believe in skill or talent, bailouts prevent asset transfers from poor stewards to potentially better stewards of capital.

    An inflation of all assets, a generalized bailout, is a bit better. It diminishes the degree to which the government picks winners and losers, and still penalizes the relatively worst choosers and stewards of assets. (More focused bailouts are often targeted towards those “in the most pain” — exactly the people who were the worst choosers and stewards of assets ex ante.)

    Debt writedowns, unsticking the liability side of the swap, should also be designed in a way to preserve incentives for good stewardship. If debts were written down systemically (e.g. a “central adjuster” modifies the value of all debt proportionately), incentives would be preserved (people who chose good assets would remain better off than people who chose poor assets).

    More individualized ways of writing down debt can preserve incentives by transferring ownership of assets disproportionately to creditors as the debt is written down. This is what currently happens in bankruptcy, but it could be done more continuously and less disruptively (see the sort of contingent capital schemes Indy mentions or my proposal for small business equity).

    Anyway, I’ve digressed. For bailouts to even plausibly be a means of desticking debt, they have to permanently rather than temporarily alter asset prices. As you suggest, short of a general inflation, this rarely happens: one time manipulation of the price of a certain asset only puts off the “reequilibration”. If this is so, as you say, it discourages new and potential better stewards of the new assets from taking them on, by creating an unhelpful post bail-out return.

    Overall I agree with you — focused asset-side bailouts (as opposed to a general inflation) — are more likely to do harm than good. They temporarily relieve the imbalance between assets and liabilities, but they invert incentives for good capital stewardship and discourage better stewards going forward.

  55. Jim Bradley — The distributional impact of central bank activity is complicated and too little studied, I think. On the one hand, economists often pretend that there are no distributional impacts, since central banks generally buy short-dated government debt with new money, minimizing the possibility of capital gains or losses. On the other hand, among anti-central-bank there is a whole theory of the advantages that accrue to the “first receivers” of new money, such as the Fed’s primary dealers. There undoubtedly are some advantages, at least in terms of transaction costs and spreads earned from intermediating open market operations. That the Fed now pays interest on reserves creates a subsidy to banks generally directly proportional to the quantity of reserves in the system (holding the rate of interest paid on reserves constant). However, one could argue that competition for “core deposits” pushes that subsidies back to savers. (I don’t buy that at all, but one could argue.)

    Anyway, people like you, and me, are suspicious. There are lots of ways the structure of central banking and the institutional context in which it is embedded might work to the benefit of some groups much more than others. You don’t have to be a black-helicopter guy like me to think that the US Fed’s connections to large banking and investment institutions had something to do with selecting what kinds of actions the Fed took and who was helped most directly when the central bank took it upon itself to “save the world” two years ago.

  56. Andy @ 24,

    First, I want to clarify that it is not my claim that central banks frequently hold down nominal wages by actively contracting. This may seem like a subtle difference, but my claim is simply that nominal wage growth exceeding productivity growth is the thing “Great Moderation” central banks are most allergic too. They do, reliably, contract when they observe rising unit labor costs, that is when they observe labor’s share of output increasing. They are very upfront about that too. But that does not imply that central banks are “the cause” of the drag we’ve observed in median wages. The Great Moderation, in my view, is mostly about central banks not having to suppress wages very much since globalization and the decline of unions broke the median wage-earner’s bargaining power. “Great Moderation” monetary policy puts a ceiling on the median worker’s share, but it was the Great Moderation precisely because wages rarely approached the ceiling that forces central banks to act disruptively.

    In my view, Great Moderation monetary interventions can be fall into two categories: “Fine tuning”, which works via cost-of-borrowing and credit and asset prices, and recessionary contraction, which works via unemployment. Since labor power has been broken, the vast majority of the Fed action has been “fine tuning”. Given suppressed median wages (mostly due to deunionizoglobalization), the Fed simply twiddles with interest rates to strike a balance between full employment and noninflation. During the Great Moderation, the marginal unit of CPI was purchased with borrowed money, not wage income, so the Fed could fight inflation without provoking unemployment. Contractions may or may not have increased real wages as you suggest. On the one hand, contractions reduced the price level relative to the price level that would have obtained absent Fed action, seeming to increase real wages. On the other hand, median wages might have been bid up some absent Fed action, so the net effect may have been to decrease the real wages of the median worker. We can’t really tell. But the marginal consumer’s purchasing power was reduced by virtue of diminished access to credit, higher borrowing costs, and lower collateral values. The sine qua non of the Great Moderation was this fact, that the marginal consumer’s behavior was sensitive to credit market changes, giving the Fed a much gentler lever than to try to push down on sticky wages.

    Prior to the Great Moderation and occasionally during (e.g. in the early 1990s), the marginal consumption dollar was wage income. Labor bargaining power was insufficiently suppressed and consumer credit insufficiently “developed” for cost-of-borrowing/credit/asset-price channels to counter wage and price pressure. During these periods, the Fed did contract to suppress wages directly. Again, your claim that this increased the median worker’s real wage relative to inaction is questionable: it’s a race between which was suppressed more, price growth or wage growth. But it did involve an increase some workers’ real wages in absolute terms. But it also involved creating a lot of unemployment. The crucial insight is in the medium term, suppressing the median workers’ bargaining power was the key to controlling inflation. Central banks could tolerate a one-time uptick in the real wages of those who continued to be employed, as long as the threat of unemployment and a reserve pool of workers meant that nominal median wages would not rise above nominal productivity, which is another way of saying that the median workers’ share of income could not be allowed to increase.

    I know this sounds like some totalizing, conspiratorial story of everything. I don’t mean it to be. But I think it should be understood as a kind of constraint — a constraint that was only occasionally binding, but nevertheless firmly enforced on those occasions. This constraint did not determine the specific path the economy took, but did limit that path to certain quadrants, and excluded certain possibilities. I agree with you that factors like competition from low-skilled labor is more responsible for the declining share of the median worker than central bank suppression. But when, occasionally, the economy did bid up median wages despite those structural factors, the Fed was always there, as a backstop to slap it down. To the degree unions still might have had bargaining power, the well-known policy of the benign, technocratic Fed — totally out of the fray of contentious labor/management dispute — helped dampen it. We all understood, there was nothing secret about it, that in order to maintain its “credibility as an inflation fighter”, the Fed would provoke a recession if necessary to prevent wage growth faster than productivity. That placed a ceiling on what kind of wage gains could be considered “reasonable”, and altered the tenor of negotiations. (I wonder if residual union bargaining power combined with an understanding that upfront wage growth would “automatically” provoke a recession played a role, along with taxes, in the shift from growth in wage to growth in benefits, a shift that has left us with some pretty big hangovers.)

    Note also that this story is consistent with the increasing positive skew of wages. It is workers who spend their wages who threaten the price level. Workers in the right-hand tail save their marginal income, investing in financial assets subject to nondisruptive fine-tuning by the central bank. The central bank didn’t “cause” wage inequality. Skills based-technical change may have had a lot to do with that, as well as increasing importance of network effects in determining wage differentials, etc. But central bank policy did restrict the universe of possible wage distributions to one that could tolerate a very long right tail but could not tolerate high real wages at the middle.

    BTW, I did a post about this story a while back, here: http://www.interfluidity.com/posts/1256656346.shtml

  57. Lord — Good point: the notion that perfect flexibility would yield perfect outcomes is an artifact of simplistic, mathy economics. It’s a convenient framework for talking about this stuff with economists, but I think that you are right that real-world economies depend as much on their frictions as on their flexibility, and that in fact we can harm economies by letting simpleminded models persuade us that it is always best to eliminate a friction. (For example, consider liquidity, or the commonplace that there would be no profits to incentivize innovation in a frictionless economy.)

    The “price” of escaping a debt includes both the interest that will be paid and the “recovery value” — how much principal will actually be paid before the debt is extinguished. Once a debt is established, we can abstract away from the income/principal distinction, and consider some kind of present value of the cash outflows we will have to support. As you say, variable rate debt tends to partially unstick the price of debt, but in the wrong direction. In the lingo, variable rates are usually “procyclical” unhelpfully rising as the burden of debt increases. Another good point!

  58. anon @ 39 — “But you’re still describing debt, aren’t you?”

    Maybe. We have to define how we’re drawing the debt/equity distinction. It’s not trivial. Are revenue bonds debt or equity? Preferred stock? Fiat currency? There are accounting definitions about these things, but the distinction is murky.

    Usually, I draw the debt/equity distinction based on whether nonpayment of an obligation triggers a disruptive bankruptcy. Even with that, there is some murk: There is such thing as nonpayment of equity, it happens when one violates priority (e.g. a dividend is paid common shareholders when a preferred dividend has not been paid), and also triggers disruptive events. But for most purposes, I think my debt/equity distinction is the useful one for policy conversations.

    In describing how financing would look in a “better world”, I won’t say whether it would “still be debt” according to my rule of thumb definition. Maybe, maybe not. The crucial point is that payoffs due would be contingent on outcomes in a way that makes outright nonpayment much less likely.

    In most economic models, there is no role for simplicity. Agents are viewed as informationally unconstrained, and their behavior is deterministic and not subject to mistakes.

    However, that is to the discredit of economic models and does not undermine your point. The reason straight debt contracts are popular means of financing assets despite their poor risk characteristics (putting aside speculative disagreement, which is also usually unmodeled) is their simplicity and their robustness to creditors’ lack of information. I think it’s possible to design better financing schemes that real humans can understand and use. (I think information technology really matters here, and that the space of possible contracts for financing real assets as opposed to refinancing financial assets has not been much explored.) But those schemes won’t perfectly reproduce the very fine grained contingent payoffs that agents would develop in a hypothetical Arrow-Debreu model with complete markets and perfect information. The real world will always have frictions and rigidities (and as Lord pointed out above, sometimes that is for the best). The point is not perfection, but to note that the status quo is far enough away from perfection that incremental improvements are called for, whether by centrally adjusting the bite of debt, decentrally streamlining debt/equity recategorization, or by developing better contracts to start with. In any case, those improvements will have to remain subject to a comprehensibility constraint. They cannot be so complex that human beings are unwilling to use them or fail to understand their implications.

  59. De K @ 30 — There are lots of ways of thinking about this.

    First, while there is optimality in models, there is not optimality in the real world, just a vast space of possible outcomes.

    Suppose that there are two paths for the economy, one with wages suppressed relative to the other.

    Does the first path have less labor supplied? Maybe, or maybe “not really”. To the degree labor markets are monopsonistic, that is to the degree that the labor supply curve is very steep, adjustments to the demand for labor will impact its price more than its quantity. It’s quite possible that multiple market wages are possible with very similar amounts of labor supplied. (This idea is at the heart of debates over minimum wages. Drawing diagonal supply and demand graphs, people with a bit of economics can’t see how pushing up the price of labor won’t reduce the quantity demanded. But if the labor supply curve is nearly vertical, as long as the effect of a minimum wage law on labor demand doesn’t push the equilibrium price below the minimum wage, the market will find a new equilibrium with a similar quantity of labor supplied at the higher wage.) In monopsonistic labor markets, lots of different wage rates are possible with a similar quantity of labor demanded. And a lot of people (left-ish economists usually) suggest a monopsonistic labor market is where the median worker lives.

    Another issue is “unemployment”. Unemployment as conventionally described is a very shifty concept. Let’s put away the whole monopsony thing and suppose that a low-median-wage path will see less labor supplied than a high-median-wage path. Does that mean higher unemployment in the low median wage path? No. If the cause of the change was fewer potential workers choosing to take a job at the low wage, that’s consistent with full employment: unemployment is defined as people willing to take a job at the going rate but who haven’t found one. So it could well be that labor force participation is suppressed in a low wage path without increasing unemployment. (This still might lead to less-than-otherwise GDP, but without unemployment, we are unlikely to call where we are an “output gap”. We’d just say that people prefer housework or leisure or whatever.)

    I think the monopsony story is more informative than the labor force participation story. But both are there in the background. And with either story, wage suppression relative to some reasonable alternative is possible without provoking measured unemployment.

  60. VJK @ 31 — I’m not sure what you mean by “the bulk of credit is *already* issued in the form of equities and bonds”. What I mean by “I think we want our economic system, including our monetary arrangements, to be based much more on equity than ‘credit’.” Is that I want much more of the financing for real assets and activities to have repayment contingent upon how those real projects work out. That’s obviously not true with bond financing. The vast majority of household financing (other than self-financing) is straight debt. Pretty much only firms are financed in part by claims contingent on anything other than disruptive default or bankruptcy, and a lot of business financing is via bonds or bank debt. So I think is lots of room to move towards more contingent forms of finance.

  61. Nick @ 41 — I think we’re mostly agreeing, or to the degree we are disagreeing it is over what would constitute “a world of difference”.

    Similarities between sticky debt prices and sticky good prices:

    1) Both represent rigidities that wouldn’t exist under the sort of complete, frictionless, markets with perfect information that would in theory yield Pareto optimal outcomes;

    2) Both can be described (for any particular referent, be it a good, service, or debt contract) in terms of two numbers, a number that actually obtains in the world (the “sticky price”) and another number, a “shadow price” describing what the price would be if there were no rigidities;

    3) In both cases, the rigidities bite. The fact that the price “sticks” rather than following the shadow price is harmful to real economic outcomes;

    4) Both are perceived by agents in the economy as a price, meaning something they have to pay in order to get something that they want;

    5) Both are potentially amenable to monetary or other policy that would prevent the sticky price and the shadow price from diverging, or that would bring them back into alignment if they do diverge. Such policy might ameliorate the harms caused by the rigidities that yield the divergence. (Or it might not — theory of the second best!)

    Differences between sticky debt prices and sticky good prices:

    1) The microeconomic specifics of how the rigidities harm real economic outcomes are very different. In the case of sticky goods prices, too few goods may be supplied or demanded, and mistaken price signals may lead to poor investment decisions over time. In the case of sticky debt prices, risk aversion due to high leverage may prevent consumption or investment, uncertainty surrounding potential defaults may make it difficult for real assets to be mobilized, frictions associated with bankruptcies may exact real tolls, unpredictable wealth transfers may blunt incentives to produce etc.

    Given this list of similarities and differences, whether “a world of difference” separates sticky goods prices and the sticky price of debt is a judgment call. Obviously, my sense is that from a macroeconomic perspective, the similarities outweigh the differences. Otherwise I’d not have made the analogy in the first place. But, if one wants to emphasize microeconomic paths rather than the effect on macroeconomic variables, one could disagree.

    This is a long way of describing your #1.

    I also think that your #2 is true. Which is just a way of saying that I think sticky debt prices (like sticky goods prices and especially sticky wages) are phenomena of the real world, not just hypotheticals.

  62. anon @ 42 — I think I’ll defer to Nick’s response @ 44 here. I don’t think a central bank could try to ameliorate the effects of sticky debt and target inflation simultaneously. It’d need another “instrument”. (such as denominating past debt in pseudocurrency that can be adjusted separately from interest rates). As things stand now, the easiest way for central banks to “unstick” debt would be to tolerate/promote a general inflation, giving up their price stability target.

  63. Nick @ 43 — I don’t disagree much when you say that I am right. At least not publicly. Between you and me, when I am not pretending to be always right on the internet I find that I am always wrong.

    I agree that there’s little inconsistency in a “first line of defense” policy against harms from sticky goods prices and sticky debt prices. Price stability helps both. I would add the caveat, again, that asset prices have to be included in the definition of price stability to keep shadow and sticky nominal debt prices from diverging.

    It’s an interesting and clever point that, since the costs of “sticky debt” is related to the variability of the burden of nominal debt and asset prices, and random walk with drift looks worse and worse over time. I agree. I’d describe this just a bit differently than an advantage of price-level targeting over inflation targeting. Instead I’d emphasize the question of how the central bank responds to deviations from its targeted path. A central bank can either accommodate past deviations, and just try to hold a target going forward, or it can precommit to a particular path. Whether that path is 0% inflation or 2% inflation, sticky debt mitigates for precommitting to a path and pushing back against any deviations, generating a mean-reverting random walk whose variance stabilizes. But inflation targeting central banks generally don’t do this, they take past noise as prologue and just try to target inflation going forward. Whether the targeted inflation rate is 0% or 2%, that generates a non-mean-reverting random walk (with or without drift), whose variance blows up over time. As Nick points out, sticky debt concerns militate towards the former sort of policy.

    (But one can still have the “lubricating effect” of inflation on sticky wages and a central bank precommitted to a fixed 2% inflation path, which policy could help manage both kinds of sticky prices.)

  64. lark writes:

    Interesting, until this:

    “I’m with you. I think unions are a poor means of supplying labor bargaining power, and wish them good riddance. I am proglobalization and free trade, or would be, if we had sense enough to subject our free trade to a balance constraint. I’ll keep your secrets. We’ll keep telling the little people that all we do is interest rates and blame whatever went wrong on Wall Street.”

    I am disappointed not only with the sentiment, but the sloppy careless ‘we are all insiders here, so let’s laugh about how we stick it to the laboring folks’ manner of expression. It is snide. It is over privileged. It lacks both seriousness and respect.

    I don’t think you realize how economics has lost credibility. Snickering at the last gasps of America’s middle class is repulsive, and deeply discredits your ideas and your blog.

  65. VJK writes:

    Steve @ 60:

    1. The firm itself makes a reasonable, one would hope, choice as to what specific IOU it wishes to use, bonds or shares, no ? We cannot force the firm prefer one instrument over the other through any means other than refusing to exchange our cash IOUs for the firm’s bonds. Besides, firms are for profit entities, they should know better !

    2. Re. households. I’d be curious to know what kind of equity-like liability issuance you have in mind for the household ? Are you concerned mainly with mortgage debts ? Should the government/taxpayers subsidize/bailout a household that has made an unreasonable decision to ‘buy’ a Mac Mansion ? How does one separate the sheep from the goats ? Etc…

  66. anon writes:

    On the one hand, you’re comparing sticky labor prices with sticky capital prices.

    On the other, you’re comparing sticky new output prices with sticky old asset prices.

    Nick 41 responded well to the second.

    Regarding the first, monetary tightening pushes up the cost of capital. Other things equal, it pushes down the cost of labor. This puts downward pressure on wages. You seem to be suggesting an “escape valve” whereby labor’s share of the adjustment is mitigated by reversing some of that increase in the cost of capital. If I borrow at 5 per cent, and the principal is written down, my cost of capital ex post falls.

    In the case of labor, there really is no way of making that kind of intertemporal adjustment between the cost of labor and the “value” of labor as an asset. So you are effectively suggesting that because the cost of capital can indeed be capitalized, labor is at an unfair disadvantage in terms of the adjustment points for its share of income and wealth.

    BTW, I understand JP Morgan’s results were partly due to writing down the value of some of its liabilities. That’s probably the kind of effect you’re not aiming for. But … ?

  67. DVD Ripper writes:

    Was very meaty. it was wonderful. let me good bold breadth of vision. it’s nice. i’ll is your regular caller. this is surely.

  68. fresno dan writes:

    Nice post.
    As a wise man once said, “If you only have a hammer, every problem is a nail.”
    The core ideology of central bankers is debt. When Summers, Bernanke, Geithner talk about credit being the “lifeblood” of the economy, they mean it. But as I have said a zillion times, debt is not wealth, and borrowing is not income.
    And for them, bankruptcy is the ultimate “f” word – unmentionable in polite society. Oh, sure, the little people can default on their little mortgages, but the big banks, i.e., big credit, can never be allowed to default. (why 100% payout on AIG debt???? because essentially the debt market and central banks are synonomous) But bankruptcy would allow prices to equilibrate to wages, restore interest rates to normal, remind people of risk, and assure that reality is considered when thinking about the return on an investment. If you can’t raise wages, than decrease prices.

  69. Nick Rowe writes:

    Steve: “I’d describe this just a bit differently than an advantage of price-level targeting over inflation targeting. Instead I’d emphasize the question of how the central bank responds to deviations from its targeted path. A central bank can either accommodate past deviations, and just try to hold a target going forward, or it can precommit to a particular path.”

    That (and what follows) is exactly how I, and most macroeconomists, would describe the crux of the issue too. We say that *is* the difference between price level path targeting and inflation targeting. Talking about price level *path* targeting allows for the possibility that the path might be upward-sloping, so you aim for 100, 102, 104, 106, etc. If you make a mistake in the second year, and hit 101 instead of 102, a price level path targeter would try to get back to the original path, while a 2% inflation targeter would shrug his shoulders and aim for 103, 105, etc. (It leaves open the question, of course, of *how quickly* one would want to get back to the original target path, but that’s a separate issue.)

    We are agreed.

    It’s just that when you say “sticky price” to a macroeconomist, he immediately sees a demand and supply curve, and a price that is stuck away from the equilibrium, and doesn’t adjust when the equilibrium moves, so excess demand or supply changes. So when you say “debt is a sticky price”, his immediate reaction is “no it ain’t!”, and you have to explain you mean something different.

    OGT writes:

    “Nick- I think I just lost whatever remaining respect I had for macroeconomics reading your comment #41.”

    That’s because you can’t have understood what I was saying, since your response does not address my point.

  70. Nick Rowe writes:

    fresno dan writes:

    “The core ideology of central bankers is debt. When Summers, Bernanke, Geithner talk about credit being the “lifeblood” of the economy, they mean it. But as I have said a zillion times, debt is not wealth, and borrowing is not income.”

    Of course central bankers (and anyone who has taken ECON 101) knows that debt is not wealth, and borrowing is not income! Jeeez!

    “And for them, bankruptcy is the ultimate “f” word – unmentionable in polite society. Oh, sure, the little people can default on their little mortgages, but the big banks, i.e., big credit, can never be allowed to default. (why 100% payout on AIG debt???? because essentially the debt market and central banks are synonomous)”

    Nonsense. They don’t believe that at all. Central banks are about money, as medium of exchange and medium of account. And using monetary policy to control aggregate demand and prices. Debt and credit matter to central banks only insofar as they affect money, and aggregate demand.

    In fact the whole point of Steve’s post is that Steve thinks central banks pay *too little* attention to debt, and that they ought to choose monetary policy with an eye towards making debt markets work better, rather than thinking of debt as a side issue. You say it’s a “Nice post”, then argue totally the other way.

  71. Nick Rowe writes:

    Steve: there is one big problem with arguing that the labour market is monopsonistic.

    Under monopsony, the value of the marginal worker to the firm (the marginal revenue product) is above the wage. That means that firms are hungry for workers: if they can find an additional worker at the prevailing wage they will definitely want to hire him, since their profits will increase if they do. The only reason they don’t hire more workers is that they can’t get any willing to work, and would have to raise the wage to attract more workers (either away from sitting at home, or away from other firms). It would look like an excess demand for labour, rather than excess demand. It would look like the exact opposite of unemployment. There would be “Help Wanted” signs at all firms, all the time.

    Now I agree that the US labour market usually looks a lot closer to this than most countries. And especially for some low-wage jobs. But I still don’t think it’s typical (though I don’t live in the US so can’t really say). And it’s presumably even less typical this last year.

    The monopoly model predicts the exact opposite. Workers are hungry for jobs, rather than firms being hungry for workers.

    Take your pick. Obviously, we see a bit of both, at different times and in different parts of the labour market. Which is more common?

  72. Nick Rowe writes:

    Ooops: change that “rather than excess demand” to “rather than excess supply”.

  73. Andy Harless writes:

    Nick @44:

    “On both these grounds, the Fed’s official objective, unless you interpret it very charitably, is incoherent.”

    Mais non.

    Given that the Fed has only one instrument, it’s just common sense (not “very charitable”) to interpret a dual mandate as a mandate to target some combination of the two objectives (e.g. alpha*x + beta*y, or perhaps alpha*x^2 + beta*y^2). The Fed is given some leeway in judgment as to the specifics of its objective function. (If one of the weights appears to be almost zero, one can then reasonably object — as some people recently have — that the Fed is ignoring part of its mandate.) Nothing incoherent about that.

    As for the other issue, short-run monetary non-neutrality may make a difference, on average, in the long run. It depends on what you believe the shape of the Phillips curve to be. It’s true that, if the short-run Phillips curve were perfectly linear (assuming that the long-run curve is vertical — i.e. long-run monetary neutrality), then there would be no way for the Fed to, on average, pursue its employment objective, so the objective would be redundant. But the Phillips curve is probably somewhat convex (certainly Phillips’ original curve was indeed a curve and not a straight line), so the Fed can maximize average employment by keeping employment as close to equilibrium as possible. Thus, for example, if the Fed needs to reduce the inflation rate, the mandate suggests that it should do so slowly. (On the other hand, you could argue that doing so quickly would have a credibility advantage, so it’s not an open and shut case, but my point is that there’s nothing inherently incoherent about the mandate.)

    Moreover, long-run monetary neutrality is not an established fact, just a consequence of certain models. In other models (of which I expect you are aware, but I won’t allow you to casually dismiss them without a mention) there are “greasing the wheels” effects, where, because of asymmetric nominal wage stickiness in individual industries, a higher average inflation rate speeds wage adjustments and thus results in higher average employment. In such models the long-run Phillips curve is not quite vertical, and money is not neutral in the long run. Accordingly, the Fed, according to such models, faces a long-run employment-inflation tradeoff which it should make according to some objective function that is consistent with the mandate.

    (As a footnote to that last paragraph, though, while I do tend to believe in “greasing the wheels” effects, it is no longer clear to me that they are relevant to the Fed’s dual mandate, because, given the risk of a deflationary trap, the target inflation rate that is optimal for a price stability objective is probably high enough to be in a range where the long-run Phillips curve is already vertical.)

  74. Nick Rowe writes:

    Suppose we give the central bank two instruments, rather than one, so they can do normal monetary policy, AND also re-write debt contracts, to either raise or lower the nominal values of all debt in response to macroeconomic circumstances.

    That has an advantage, in principle, because with 2 instruments the Bank can hit 2 targets. But by God, it would be incredibly controversial politically. It would be like giving the Bank the power to re-write contracts.

    Even with existing monetary policy this very political element exists. Debtors like unanticipated inflation; creditors like unanticipated deflation. (I mean a higher or lower price level than was anticipated at the time the debt contract was undertaken). But an agreed-on inflation target (or price level path target) removes the central bank’s discretion on this issue, and replaces it with a quasi-constitutional rule. And I really like that aspect of inflation targeting. It makes the central bank like a monetary supreme court, which can interpret the rules (use it’s discretion on how best to hit the inflation target) but cannot just re-write the rules or make them up whenever it wants. It’s accountable.

    Under discretion, debtors usually win the political fight. They are more economically active and powerful than creditors. Granny doesn’t understand why her bonds aren’t worth as much, and can’t fight it politically anyway.

    The nearest analogy, in the US, is when gold clauses in debt contracts were abrogated in the 1930’s.

    There is definitely a risk that if the value of debts were determined by discretion rather than rules, this would worsen the functioning of debt markets.

  75. Nick Rowe writes:

    Andy: Under your interpretation it is indeed coherent. But that is precisely what I would call a “charitable” interpretation! Your interpretation of the Fed’s mandate (if y is unemployment) calls for it to aim for *stable* unemployment (minimise y^2 or variance of unemployment). But IIRC, the Fed’s mandate calls for *low* unemployment.

    What you have just described is actually very close to the Bank of Canada Act, which calls for (quoting from memory, and probably not quite right) the Bank of Canada “..to mitigate, to the extent that it can, *fluctuations* in employment (and other real stuff, plus price stability)..”. Pre-dated New Keynesianism by about 60 years!

    Another simple example. Suppose the LRPC is vertical, and the SRPC linear. Then suppose the natural rate increased, so the LRPC shifted. Should the Fed choose higher inflation?

  76. OGT writes:

    Fresno Dan- I agree with Nick. The problem is that CB’s and their economic models don’t pay enough attention to debt. In fact most macro models simply assume that debt doesn’t exist or at least that the our financial system is so perfectly functioning as to be invisible. Obviously that makes modeling easier, but I agree with Waldman that it’s a mistake.

    To quote from the Alan Taylor, Moritz Schularick paper I linked to above:

    In this paper we study the behavior of money, credit, and macroeconomic indicators
    over the long run based on a newly constructed historical dataset for 14 developed countries over the years 1870– 2008, utilizing the data to study rare events associated with financial crisis episodes. We present new evidence that leverage in the financial sector has increased strongly in the second half of the twentieth century as shown by a decoupling of money and credit aggregates, and we also find a decline in safe assets on banks’ balance sheets. We show for the first time how monetary policy responses to financial crises have been more aggressive post-1945, but how despite these policies the output costs of crises have remained large. Importantly, we demonstrate that credit growth is a powerful predictor of financial crises, suggesting that such crises are “credit booms gone wrong” and that policymakers ignore credit at their peril. It is only with the long-run comparative data assembled for this paper that these patterns can be seen clearly.

    Nick- Perhaps I didn’t follow your comment, but Waldman’s comment in response at #61 did get to the main objection I was trying to make. Also that the nominal rigidity of nominal debt contracts in relation to assets as well as the pro-cyclicality of debt and asset prices is a a significant problem, as such I found your chosen example of a car with a one year loan to be less apt in describing the more common usage of debt or its potential rigidity problems.

    Then I threw in some stuff about the difference between how most economist’s decribe the workings of monetary policy as presently practiced and how it appears to work in practice.

  77. […] Waldman thinks that an emphasis by macroeconomists on containing “sticky prices” has led to an […]

  78. shoreb45 writes:

    supposedly there is speculation that John Paulson’s fund is being forced to liquidate positions because of their massive holdings in BAC, C, and GLD, and this is adding to the pressure on gold:

    http://www.goldalert.com/gold-price-blog.php#paulson_rumor

  79. Nick Rowe writes:

    OGT writes:

    “Fresno Dan- I agree with Nick. The problem is that CB’s and their economic models don’t pay enough attention to debt. In fact most macro models simply assume that debt doesn’t exist or at least that the our financial system is so perfectly functioning as to be invisible. Obviously that makes modeling easier, but I agree with Waldman that it’s a mistake.”

    Exactly!

    But just to be picky, there are two sorts of mistakes:

    1. When debt and credit markets stop working, that affects aggregate demand, and may also affect the degree to which monetary policy affects aggregate demand. Most macroeconomists have realised this — now.
    2. Maybe conducting monetary policy differently, with more of an eye to making the ex post returns to debt closer to what borrower and lender would have wanted ex ante in an idealised world, where they could see the future, might also be desirable.

    That second one is Steve’s point. It wouldn’t be fair to say that macroeconomists have ignored this completely. But it’s (usually) not in the models, and it’s an afterthought, if it is thought of at all, when it comes to the objectives of monetary policy, and then only in a very narrow way, namely the effect of unanticipated inflation or deflation.

    “Nick- Perhaps I didn’t follow your comment, but Waldman’s comment in response at #61 did get to the main objection I was trying to make. Also that the nominal rigidity of nominal debt contracts in relation to assets as well as the pro-cyclicality of debt and asset prices is a a significant problem, as such I found your chosen example of a car with a one year loan to be less apt in describing the more common usage of debt or its potential rigidity problems.”

    OK. Maybe it’s just that macroeconomists have a more precise (narrow?) view of what we mean by “sticky price”, and “price”. For example, if I buy a beer with a $1 lottery ticket, and then the lottery ticket wins $1 million, we still say the price of the beer was $1, and we don’t say that the price was too high. And our definition of “price” matters, because it’s what matters for getting supply and demand into equilibrium.

    No worries.

  80. Andy Harless writes:

    Nick,

    I think you misunderstood my comment. (I should have left out the y^2 part because it was not central to my argument.) The Fed’s mandate is for “maximum employment”, and my main point was that, provided that the short-run Phillips curve is convex, maximizing average employment implies keeping the unemployment rate close to an optimum. Given that the old-fashioned empirical Phillips curve appeared quite obviously convex, I think the burden of proof would be on anyone who wanted to argue that the short-run Phillips curve was linear.

  81. Jim Bradley writes:

    Steve –

    I don’t think it’s quite that ambiguous … In my view, The Federal Reserve, from the perspective of Congress, exists to finance the debt (more than the unimpeded market would take with an inelastic currency – like gold – which maintains fiscal discipline) of the U.S. government through the international banking system. The Federal Reserve, from the perspective of elite finance, exists to extract a significant portion of value from laborers and the less-politically connected, and to control the U.S. government into paths that benefit the elite.

    I think it’s more than a legitimate worry, as our democracy is hollowed out, that these same jokers will turn our country into a totalitarian state to “maintain their gains” as the corrective forces of the market get larger.

    At least, that paradigm has played out with recent events.

    The potential for the next step is also a problem. To maintain liquidity and store the “gains” by using financial assets, there must exist a currency and a debt structure that is relatively sound. If the U.S. becomes precarious, the plan would have to be to create a global currency (or a currency system large enough to absorb the same value as the massive dollar flow per day: roughly 2 trillion at present) … perhaps it will be backed by “energy taxes”? Every possibility appears to indicate some sort of multinational global governance.

    However, since the Euro is falling down, the multinational paradigm is failing. That probably means some event will have to occur to rally the world around the new centralized system. I hope it doesn’t come to that … but again, if recent events show that we have yielded national sovereignty to elite finance, that is the logical outcome. It could be that this outcome is as sure as the logical outcome of having a “Federal Reserve” causes the problems we have today.

  82. winterspeak writes:

    Nick Rowe said “But that $100 I owe you is a transfer payment. It’s like a welfare payment or social security check. The decision it affected has already been made, last year. It cannot any longer create excess demand or supply, though it could have done last year, if it had been sticky last year. What it does do is transfer resources from me to you. It affects the distribution of wealth.”

    $100 in debt is not a transfer payment. It is not a transfer payment when its made, and it isn’t a transfer payment when it’s paid back. It only becomes a transfer payment if it’s defaulted on. And budget constraints, which include debt (outstanding, money good, and written down) absolutely impact forward-looking decisions.

  83. RueTheDay writes:

    @ Nick Rowe and Steve – While it may not be theoretically correct to call a debt service payment a “price” in the same sense as one calls the current wage rate the price of labor, it does reflect a cost of production in the same way that labor and raw materials do. It is in this sense that the fixed, nominal interest payments of previously entered into debt contracts introduce a rigidity into the system.

    In Chapter 21 of Irving Fisher’s, “The Theory Of Interest”, he introduced six principles that were required to make the equilibrium interest rate theory work. The first two were:

    1. The market must be cleared—and cleared with respect to every interval of time.
    2. The debts must be paid.

    It was these first two principles that Fisher later rejected in his 1933, “The Debt-Deflation Theory Of Great Depressions”.

    Specifically, it is the error behind principle number two that refutes Nick’s claim that debt service payments simply represent a transfer payment between debtor and creditor with no aggregate effect.

  84. RueTheDay writes:

    It appears Paul Krugman has now jumped head first into the MMT debate, via his back and forth with Jamie Galbraith:

    http://krugman.blogs.nytimes.com/2010/07/17/i-would-do-anything-for-stimulus-but-i-wont-do-that-wonkish/

    http://krugman.blogs.nytimes.com/2010/07/17/more-on-deficit-limits/

  85. winterspeak writes:

    RTD: Yes, and I hope this confirms that Paul is not MMT once and for all.

  86. Steve Roth writes:

    Just to say thanks, I’ve been wondering about this issue forever. The bellwether of inflation — what leads the fed to clamp down — is rising wages. To rank amateurs like myself, it looks decidedly like the game is heavily rigged to favor capital over labor.

    The result has indeed been moderation — notably decided moderation in lower- and middle-class incomes. And decidedly otherwise for returns on capital.

    I think the simplest solution is a tax system that (combining state, local and fed taxes) actually *is* progressive, with the extra money going to broadly based transfers, notably the Earned Income Tax Credit. (While also increasing that credit’s “saliency” by delivering it in weekly paychecks.

    Employers should *love* this idea.

  87. Tom C writes:

    Throughout this post and the comments, I see people trying to apply economic theory to target the question about the stickiest price. That raises a question for me – is it possible to reach a point of such economic upheaval that traditional (or even more modern) economic theory just doesn’t cut it. In other words, there’s talk of wage expectations and soaring debt, but at what point does the imbalance become so great that the solution is outside our normal economic thought sphere?

  88. Morgan_03 writes:

    Steve,

    Great post, I need to further digest while trying to finish the “Girl Who Kicked The Hornets Nest.” My feelings on debt jubilee/restructuring have changed 180 degrees over the past year.