Interesting links

Here’s a big backlog of things I wanted to blog about, but don’t have time to cover in depth:

1.  Jim Hamilton discusses a fascinating study on the effects of colonialism by Feyrer and Sacerdote .

2.  David Henderson asks about the balanced budget multiplier being one.  Why would workers produce more for no increase in after tax income?  Isn’t the Keynesian answer that workers have no say in how much they work?

3.  Bryan Caplan points out that they blew it on the payroll tax cut.

4.  Arnold Kling and Russ Roberts make AD seem more mysterious than it really is.  When both output and inflation rise, AD has risen.  When output rises and inflation falls, AS has risen.  (Oops, replace “inflation” with NGDP growth.)

5.  Bryan Caplan nails it: “When “they” take over, we’ll be them.”  I recall reading that by 2050 the US government will consider America to be “50% white.”  But at the same time 71% will self-identify as “white.”  I think it will be much higher than 71%.

6.  Moral superstars in Britain.  Tyler Cowen is right, in terms of morality they are better people than you and I.

7.  Karl Smith points out (correctly) that just because progressives are hypocrites, doesn’t mean they are wrong:

I suggest that a rich person can consistently favor taxes on the rich without volunteering to pay such taxes him or herself. The argument is simply that the world in which every rich person pays is preferable to me to the world in which no rich person pays which is preferable to me to the world in which only I pay.

Let he who has (anonymously) donated a kidney to a stranger cast the first stone.

8.  Arnold Kling exposes the myth that deregulation led to the banking crisis.

9.  Matt Yglesias gets it:  NGDP is the actual “thing.”

1o.  Bill Woolsey shows that it doesn’t matter all that much which version of NGDP you target.

11.  David Beckworth refutes Joe Weisenthall.

12.  The wisdom of Nick Rowe:

The interest rate language they currently use does not contain negative numbers. And it is an ambiguous language when it comes to whether an increase in nominal interest rates means a tightening or a loosening of monetary policy. . . . It might be pushing things only a little too far if I said that economics is a sub-field within philosophy of language.

13.  Jesse, don’t leave yet—there’s still a sliver of hope.

14.  I’ve noticed that all the “Don’t be naive Sumner, the Fed holds short term rates below long term rates to help the big banks” comments have mysteriously stopped with Operation Twist.

I will be traveling, and hence may be slow in responding to comments.


Tags:

 
 
 

62 Responses to “Interesting links”

  1. Gravatar of Adam Adam
    6. November 2011 at 10:07

    It’s true, NGDP is the actual thing. But even so, it’s created using survey data, This obviously has various degrees of robustness – often the survey for one sector is ropey and off-model adjustments are made to it. You only have to look at the variation between the first published results and the corrections to see that it contains in it huge amounts of variation.

    The grey market Is interesting too since NGDP surveys struggle to take it into account, traditional industries can report a dip in output when asked, and that would show as an Ngdp dip, but actually the grey market has taken the slack.

  2. Gravatar of Policy Wank Policy Wank
    6. November 2011 at 11:21

    Based on her diavlog with Karl Smith I would say that Kelly Evans is still fairly confused about NGDP targeting. Unfortunately, Karl does not do the best job of understanding what her points of confusion are and clearly addressing them. Hopefully Scott can do a better job.

    http://bloggingheads.tv/diavlogs/39713

  3. Gravatar of Peter N Peter N
    6. November 2011 at 11:57

    I’m quite unimpressed with the rather obtuse Mr. Kling. The comments to his foolishness are rather better.

    He finds:

    “1. The conditions that are necessary in order to be able to interpret market demand in terms of a representative consumer are unrealistically stringent.

    2. The conditions under which market demand absolutely must slope down if individual demand curves slope down are unrealistically stringent.”

    to be “uninteresting”. Well I find his criticism of Keen uninformed and more than a bit disingenuous. See, I’ve refuted him. So easy!

    Keen is restating results that are widely accepted and then ignored by nonspecialists. A lot of what he says is based on
    the Sonnenschein–Mantel–Debreu theorem.

    From Wikipedia:

    “Formally, the theorem states that the Walrasian aggregate excess demand function inherits only certain properties of individual excess demand functions:

    Continuity

    Homogeneity of degree zero,

    Walras’ law, and a boundary condition assuring that as prices approach zero demand becomes large.

    These inherited properties are not sufficient to guarantee that the aggregate excess demand functions obey the weak axiom of revealed preference : this [sic] aggregate demand functions can have “any shape”, which means that in perfect competition model, it is impossible to deduce from a maximizing behavior of households and firms the shape of their supplies and demands. This has lots of consequences for the microeconomic field. Most notably, the uniqueness and the stability of the equilibrium is not guaranteed : it may have more than one root – more than one price vector at which excess demand is zero (the standard definition of equilibrium in this context) and no general process directing toward any equilibrium point (such as the famous Walrasian “tâtonnement” process) can be deduced from the assumptions.

    But, as Rizvy has explained, the range of implications are not here limited: “There are problems with establishing general results on uniqueness (Ingrao and Israel 1990,chap. 11; Kehoe 1985, 1991; Mas-Colell 1991), stability (Sonnenschein 1973; Ingrao and Israel 1990, chap. 12; Rizvi 1990, 94–144), comparative statics (Kehoe 1985; Nachbar 2002, 2004), econometric identification (Stoker 1984a, 1984b), microfoundations of macroeconomics (Kirman 1992; Rizvi 1994b), and the foundations of imperfectly competitive general equilibrium (Roberts and Sonnenschein 1977; Grodal 1996). Subfields of economics that relied on well-behaved aggregate excess demand for much of their theoretical development, such as international economics, were also left in the lurch (Kemp and Shimomura 2002).”[1]

    Occasionally the Sonnenschein–Mantel–Debreu theorem is referred to as the “Anything Goes Theorem”.[1]”

    Clearly this is “uninteresting”. I wonder why all these economists have been wasting their time on it?

    But there’s more:

    “3. The conditions are under which market demand is likely to slope down if individual demand curves slope down are unrealistically stringent.”

    ” Proposition (3) would be interesting, but it is not true. Keen uses propositions (1) and (2) to dance a sort of intellectual strip tease that makes you think he is going to reveal a case for (3), but he never goes that far.”

    Perhaps a short course in remedial reading would help.

    Keen quotes Mas-Colell’s Microeconomic Theory [a widely used textbook]:

    “If there is a normative representative consumer, the preferences of this consumer have welfare significance and the aggregate demand function can be used to make welfare judgments by means of the techniques [used for individual consumers]. In doing so, however, IT SHOULD NEVER BE FORGOTTEN THAT A GIVEN WEALTH DISTRIBUTION RULE [imposed by the 'benevolent central authority'] IS BEING ADHERED TO, AND THAT THE ‘LEVEL OF WEALTH’ SHOULD ALWAYS BE UNDERSTOOD AS THE ‘OPTIMALLY DISTRIBUTED LEVEL OF WEALTH [caps for italics].”

    So all you need to do is assume everybody has the same preferences and the government redistributes wealth optimally. Anybody have a problem with that?

    As for regulations, I may deal with that later.

  4. Gravatar of Greg Ransom Greg Ransom
    6. November 2011 at 11:58

    Scott, you’d like the last 30 or 40 pages of Hayek’s _The Pure Theory of Capital_ were he destroys the idea that interest rates always and necessarily indicate either the “looseness” of monetary policy or even the demand for investment and the supply of savings.

    You credit Friedman with this insight. Others credit Keynes. Hayek has a much firmer grasp of the details and fundamentals of matter than either.

  5. Gravatar of Turner Turner
    6. November 2011 at 16:42

    ‘Arnold Kling exposes the myth that deregulation led to the banking crisis.’

    It is simple reality that deregulation caused the banking crisis. You’d have to blind or ignorant (in the true sense of the word, not the all purpose insult sense) to see that.

    All I see from Kling is: banks were regulated, so it was a ‘government failure’, not a ‘market failure’, which only makes sense in neoclassical fantasy land of governments versus markets. If you stick by that line, the real culprit is clearly limited liability laws. But I don’t see many ‘libertarians’ calling for an end to those.

  6. Gravatar of Peter N Peter N
    6. November 2011 at 19:10

    Regulation – “It’s not the meat, it’s the motion.”

    Counting regulations is like counting Federal reserve notes. I have 1000 of them. If they are $100 bills I’m doing OK. $1, not so much.

    Greenspan kept interest rates low. Investors wanted better returns but still on investment grade instruments. The banks obliged. Nothing wrong with that, provided it’s on the up and up.

    However, the new wild west was without a Judge Roy Bean.

    “4 Derivatives had become a uniquely unregulated financial instrument. They are exempt from all oversight, counter-party disclosure, exchange listing requirements, state insurance supervision and, most important, reserve requirements. This allowed AIG to write $3 trillion in derivatives while reserving precisely zero dollars against future claims.

    5 The Securities and Exchange Commission changed the leverage rules for just five Wall Street banks in 2004. The “Bear Stearns exemption” replaced the 1977 net capitalization rule’s 12-to-1 leverage limit. In its place, it allowed unlimited leverage for Goldman Sachs, Morgan Stanley, Merrill Lynch, Lehman Brothers and Bear Stearns. These banks ramped leverage to 20-, 30-, even 40-to-1. Extreme leverage leaves very little room for error..”

    http://www.washingtonpost.com/business/what-caused-the-financial-crisis-the-big-lie-goes-viral/2011/10/31/gIQAXlSOqM_story_1.html

    “While he was at the Treasury Department, Summers’ enthusiasm for financial deregulation conflicted with the views of Brooksley Born, who became Chair of the Commodity Futures Trading Commission in 1996. Born had extensive previous experience working as a lawyer in the derivatives area and was concerned about the lack of oversight of this multi-trillion dollar financial market. However, Summers collaborated with Alan Greenspan, Robert Rubin, and Arthur Levitt to block Born’s efforts to initiate discussion of regulating derivatives. According to New York Times report Timothy O’Brien “They were all part of a very concerted effort to shut her up and to shut her down. And they did, in fact, shut her up and shut her down. Bob Rubin is not a guy who likes confrontation. He’s confrontation-averse. But he understands that you need someone in there who can swing a heavy axe, and that person was Larry Summers. He was the enforcer.” [20]

    http://sourcewatch.org/index.php?title=Larry_Summers#Opposition_to_Regulation_of_Derivatives

    So far it’s likely that this would have been no worse than another S&L crisis, but Wall street discovered the magic of derivatives, particularly the Repo market.

    Assume 2 markets one where money has a velocity of around 3 call it SM and another where it’s more like 100, call it FM. Barriers to entry prevent merger, but it is possible to take a security from SM and exploit it in FM, multiplying rents by the difference in velocity. Sort of like check kiting. You can buy a security in SM, Repo it, and use the proceeds to buy another. You take a cut of the interest payments from every mortgage. Even 2 % ROI is nice multiplied by 30.

    You can cover some of your exposure by getting a derivative from the clueless turkeys at AIG (who wrote $3 trillion of derivatives). Or maybe you create a second (or third, or fourth) order derivative.

    But surely even though there were over $100 trillion of derivatives outstanding, it was all hedged. Mostly true, but hedges depend on counter-parties. You have to be able to deal with a counter-party’s right to additional collateral from you (a kind of margin call) which can force you to raise capital by finding new investors (who, you can be sure will be numerous and generous, as the European banks are now discovering) or dumping assets in a fire sale, and deal with a counter-party’s inability to meet such a call from you, turning your hedge into a position rapidly heading south. Quite possibly both at the same time. Lehman, and Bear couldn’t. Systemic risk, I think it’s called.

    Finally there’s always the problem of your unhappy customers, from whom you’ve profited handsomely, taking legal action alleging things like fraud, failure to disclose material risks and abuse of fiduciary relationship. I’m still waiting for criminal conspiracy. It’s in the wings.

    Is BOA solvent or will they choke on Countrywide? Will they be able to put Countrywide into bankruptcy without dragging in the parent holding company? Stay tuned.

    There’s an old saying – A fish rots from the head.

    There’s more in the 2 links.

  7. Gravatar of johnleemk johnleemk
    6. November 2011 at 21:33

    The Dartmouth International Students Association sponsored a dinner discussion with Sacerdote a couple years ago where he talked about his colonial islands research. Really fascinating stuff, which I suppose is how I wound up double majoring in econ (Feyrer actually signed my major card) and history.

    BTW the link on Hamilton’s blog to their paper is broken; the correct link is http://www.dartmouth.edu/~jfeyrer/islands_2007_09.pdf

  8. Gravatar of Paul Andrews Paul Andrews
    6. November 2011 at 23:21

    “NGDP is the actual “thing.””

    Empty rhetoric I am afraid, and highly symbolic of economists’ tendency as a group to hyper-caricature highly complex phenomena, just so they can “understand” them, then pretend their caricature is the real thing.

    What matters is that people are busy producing goods and services that other people want, in exchange for goods and services that they want. This is not reducible to a number. RGDP is highly imperfect and NGDP is worse, as an attempted measure of economic health.

  9. Gravatar of StatsGuy StatsGuy
    7. November 2011 at 07:05

    Not impressed by the Fayer/Acerdote argument. The first stage fit is weak, largely driven by a few outliers, and that means an unstable IV estimator which is heavily dependent on assumptions.

    That doesn’t even address the question of no direct effect/alternate causal path between the IV and the final dependent variable, which Hamilton addresses.

    Reminds me of the malaria paper. BTW, even if you believe the argument in these papers (settlers brought institutions), we still do not see effective separation between institutions and other ‘pieces’ of society they brought along, like trade networks, technology, business practices, etc.

  10. Gravatar of Morgan Warstler Morgan Warstler
    7. November 2011 at 10:01

    It is weird how liberals come to terms with Greece working out the way I said it would:

    http://thinkprogress.org/yglesias/2011/11/07/362376/greeks-going-grand-coalition/

    Mundell really is a genius. In one fell swoop, he’s fixing almost everything I didn’t like about Europe.

  11. Gravatar of Rob Rob
    7. November 2011 at 10:04

    FWIW, I had the same take on the Fayer/Acerdote argument just glancing at the scatter plot and considering Hamilton’s alternative explanation, which is more intuitive. In any event parsing the effects of institutions (at a very high level of aggregation) and resource endowments based on 200 year old wind data seems inherently risky!

    There’s a comment posted that links to a paper by Iyer that looks to me to be quite interesting on this topic. Iyer examines Indian states directly and indirectly controlled by the British and concludes that states indirectly controlled have better post-colonial access to things like schools, health centers, and roads. http://www.mitpressjournals.org/doi/pdf/10.1162/REST_a_00023

    The hypotheses aren’t mutually exclusive of course. It could be that colonialization was good for islands but bad for regions in co-evolution and/or competition. There’s more questions here than clues, it seems to me.

  12. Gravatar of Morgan Warstler Morgan Warstler
    7. November 2011 at 10:40

    Hope for America…

    See DeKrugman’s column today get ass-slapped by the wise crowd of Hacker News.

    http://news.ycombinator.com/item?id=3206138

  13. Gravatar of Silas Barta Silas Barta
    7. November 2011 at 14:28

    I don’t get Yglesias’s point. (Well, I get it, it’s just stupid.) Yes, NGDP is the easiest quantity to concretely define and measure. The problem is that it is irrelevant as a thing we care about beyond its instrumental value for what we do care about (and is harder to measure).

    Despite the impression you may have gotten from Scott_Sumner, no one actually cares how many times a dollar is transferred to another person for a purchase in a given time period! Nor do businesses decide how much of their savings to blow based on how much they expect NGDP to change.

    What we care about is how well the economy satisfies our wants. You can make NGDP (or expected NGDP) anything you want and still fail by that metric!

    So the fact that NGDP is easier to measure is irrelevant, and he’s answering the wrong question.

  14. Gravatar of ssumner ssumner
    7. November 2011 at 20:20

    Adam, The revisions to NGDI are much smaller, so that’s probably the best aggregate to target. Of course NGDI is exactly equal to NGDP, but the initial estimates can differ, and reported NGDI is probably the best estimate of NGDP.

    Policy Wank, You’ll have to let me know, as I can’t bear watching myself.

    Peter, I agree that his stuff on AD leaves something to be desired.

    Greg, Thanks for that info. I didn’t mean Friedman was the first to notice this problem, just that he stated is forcefully in regard to Japan, and also the US in the 1930s.

    Turner, you said;

    “All I see from Kling is: banks were regulated, so it was a ‘government failure’, not a ‘market failure’, which only makes sense in neoclassical fantasy land of governments versus markets.”

    Then you need to reread his post. He agreed there was deregulation, but he pointed out that the deregulation occurred in areas unrelated to the banking crisis. Things like branching laws, or interest rate ceilings.

    Peter, I don’t see derivatives as being the main problem. For me the mains problems were:

    1. Loans from small banks to developers. That explains a big part of the losses to FDIC.

    2. Sub-prime mortgages.

    3. Fannie and Freddie. The bailout there may exceed $100 billion.

    4. Big bank purchases of MBSs.

    None of those are that closely related to deregulation, except maybe sub-primes–but that sort of deregulation was supported by those on the left who oppose “deregulation.” So the mantra on the left that “deregulation” was the problem is somewhat misleading.

    Thanks Johnleemk.

    Paul, Who ever said NGDP was used to measure economic health? Otherwise Zimbabwe would be the healthiest economy on Earth.

    Statsguy, Thanks for that appraisal. I admit to not talking a close look. I just thought it was a neat natural experiment. But then the best experiments often have disappointingly low statistical significance.

    Morgan, I saw him interviewed recently–he didn’t look like someone whose grand dreams were all working out as planned.

    Rob, That’s an interesting finding in India–is it statistically significant at a high level?

    Silas, You said;

    “Despite the impression you may have gotten from Scott_Sumner, no one actually cares how many times a dollar is transferred to another person for a purchase in a given time period!”

    You are continuing your near perfect record of misquoting me in every single comment. I never said velocity was worth paying attention to.

  15. Gravatar of Paul Andrews Paul Andrews
    7. November 2011 at 21:28

    @Scott:

    At http://www.nationalaffairs.com/publications/detail/re-targeting-the-fed you say: “This suggests that NGDP is useful not only as a predictor and indicator of trouble, but as a target for monetary policy.”

    Above you say: “Who ever said NGDP was used to measure economic health?”

    If it is not a measure of economic health then it cannot be an indicator of economic trouble.

  16. Gravatar of Paul Andrews Paul Andrews
    7. November 2011 at 21:42

    @Scott:

    You say above: “I never said velocity was worth paying attention to.” (in reference to “how many times a dollar is transferred to another person for a purchase in a given time period”).

    The number of times a dollar is transferred to another person for a purchase in a given time period *is* NGDP. Velocity is this number divided by the number of dollars available for transfer.

  17. Gravatar of W. Peden W. Peden
    7. November 2011 at 22:59

    Paul Andrews,

    Strictly speaking, “velocity”* (horrible name) is how many times the TOTAL money stock is exchanged for FINAL GOODS within a given time period. Saying NGDP is velocity is like saying energy is the speed of light squared: it is just one half of the left side of the equation.

    Someone can be interested in NGDP and not case about dividing it by the money stock in order to calculate velocity.

    * In the way we calculate it, NOT as it is defined in MV = PT. NGDP is just a proxy for PT, which includes things like asset transactions and intermediate transactions.

    Silas,

    In what way is your criticism different from “NGDP targeting does not solve all the problems of the world, therefore it is worthless”? Why should a monetary system be expected to resolve the problem of good vs. bad transactions?

  18. Gravatar of Paul Andrews Paul Andrews
    7. November 2011 at 23:21

    W. Peden,

    You said: “Saying NGDP is velocity is like saying energy is the speed of light squared”

    Who said NGDP is velocity?

  19. Gravatar of W. Peden W. Peden
    7. November 2011 at 23:52

    “The number of times a dollar is transferred to another person for a purchase in a given time period *is* NGDP.”

    The subject of that sentence is velocity. NGDP is the money stock multiplied by the (average) number of times a given unit of currency is exchanged for a final good.

    If we’re interested in controlling NGDP, then we are interested in keeping the money supply in line with the demand to hold money. For example, this would have implied a much tighter monetary policy in the 1970s, when NGDP shot up. A productivity norm, where NGDP is targeted at the trend rate of total factor productivity, would imply a deflationary monetary policy rather than the inflationary monetary policy of the last 100 years in the US.

    Now, no NGDP targeting regime is going to solve problems caused by misallocations of resources. That is Silas’s big point. Whether that fact is relevant or about as significant as the observation that athlete’s foot cream doesn’t cure acne, I leave to you to decide…

  20. Gravatar of Paul Andrews Paul Andrews
    8. November 2011 at 01:03

    @W. Peden,

    You implied that I said NGDP is velocity.

    What I said was “The number of times a dollar is transferred to another person for a purchase in a given time period *is* NGDP.”

    You now say that the subject of that sentence is velocity, in defending your implication that I said NGDP is velocity.

    Perhaps there is some semantic confusion here.

    When I use Silas’ phrase “The number of times a dollar is transferred to another person for a purchase in a given time period”, I mean (and I presume Silas means) “The number of times anyone transfers any dollar to another person for a purchase in a given time period” – i.e. the total number of dollars spent for purchases in the economy.

    “Now, no NGDP targeting regime is going to solve problems caused by misallocations of resources. That is Silas’s big point. Whether that fact is relevant or about as significant as the observation that athlete’s foot cream doesn’t cure acne, I leave to you to decide…”

    If my doctor is convinced that athlete’s foot cream can cure all my other illnesses despite its inefficacy with respect to acne, and doesn’t think acne a huge cause for concern, and I know that the cream may worsen the acne, and the doctor’s only diagnostic tool is to check that the same dose of cream has been applied each day, and to disregard all other symptoms, then yes I think it is relevant.

  21. Gravatar of W. Peden W. Peden
    8. November 2011 at 03:28

    It does look like the disagreement here is semantic rather than real. The point to take away here, I think, is that NGDP targeting means a desire to keep the money supply from running all over the place + attention to the demand to hold money.

    “If my doctor is convinced that athlete’s foot cream can cure all my other illnesses despite its inefficacy with respect to acne, and doesn’t think acne a huge cause for concern, and I know that the cream may worsen the acne, and the doctor’s only diagnostic tool is to check that the same dose of cream has been applied each day, and to disregard all other symptoms, then yes I think it is relevant.”

    Can you name such a doctor? Scott Sumner has talked before about how many of the United States’s problems are structural e.g. extremely rapid rises in real minimum wages.

    It seems to me that, if someone has nothing to say about NGDP targeting as a monetary system, they resort to pointing out the obvious fact that athlete’s foot doesn’t cure acne.

    Of course, if you want to make the point that the cream might make things WORSE, then do so.

  22. Gravatar of Morgan Warstler Morgan Warstler
    8. November 2011 at 04:50

    Here’s the best read of the day:

    http://blogs.the-american-interest.com/wrm/2011/11/06/occupy-blue-wall-street/

  23. Gravatar of Rob Rob
    8. November 2011 at 06:02

    “Rob, That’s an interesting finding in India–is it statistically significant at a high level?”

    You’re going to be sorry you asked. :)

    Hmmm…..Took a closer look myself. The short answer is kinda-sorta – many of the analyses are significant at around 10% for roads, schools, and health centers. As these things go, it’s a complicated analysis and the result depends a lot on controls, measures, and time periods. Perhaps an interesting finding in and of itself is that the differences get washed out over time.

    I agree that the search for an exogenous variable to stratify the sample and get independent estimates of resource endowments and institutional effects is what’s interesting in these papers. Feyrer/Sacerdote use wind. Iyer’s is a neat experiment too, in my opinion.

    The problem with the India analysis is that the colonials apparently chose areas with agricultural potential to annex, as might be expected. Iyers innovative “instrumental variable” therefore is to look at those states impacted by the “Doctrine of Lapse.” A British Lord implemented this doctrine, under which the British annexed states where a ruler died with no natural heir (the policy was designed to not recognize adopted heirs). Thus, the states under indirect rule (mostly foreign policy but not internal administration) were randomly selected based on royal mating habits instead of on agricultural productivity. Insert quip about the amazing powers of basmati rice here……

    By the way, the paper also speculates why hereditary kings were better managers than British landlords and gives somewhat of an agency argument. “The key difference is that kings were explicitly subject to being removed in cases of gross mis-rule, while landlords did not have this institutional con- straint. (709).”

    Double Hmmmm….. I’d expect more of a red queen argument with a nationalistic flavor – the kings need to work harder just to stand still – but whatever.

  24. Gravatar of Understanding the Balanced-Budget Multiplier Theorem « Uneasy Money Understanding the Balanced-Budget Multiplier Theorem « Uneasy Money
    8. November 2011 at 07:16

    [...] Sumner recently linked to David Henderson who cited the following comment by Professor T. Norman Van Cott of Ball State [...]

  25. Gravatar of Colonialism, Tax Cuts, Redistributionists, and Other Links | John Goodman’s Health Policy Blog | NCPA.org Colonialism, Tax Cuts, Redistributionists, and Other Links | John Goodman's Health Policy Blog | NCPA.org
    8. November 2011 at 07:26

    [...] These are courtesy of Scott Sumner: [...]

  26. Gravatar of Eric Morey Eric Morey
    8. November 2011 at 07:37

    Scott,
    I’m a bit baffled by your view of the main problems.

    “I don’t see derivatives as being the main problem. For me the mains problems were:
    2. Sub-prime mortgages.”

    The vast majority of Sub-prime mortgages were packaged into MBS’s and CDO’s in multiple levels i.e. derivatives allowing for increased financial leverage for non-bank firms. Limits on leverage and derivatives would have reduced the amount of Sub-prime mortgage dollars lent. Do you not agree? Why?

    “3. Fannie and Freddie. The bailout there may exceed $100 billion.”
    Didn’t our government take ownership of the GSE’s? There is value that the taxpayers now receive from the cash flow of their operations. Is this factored into your statement about Fannie and Freddie?

    I don’t understand why the GSE where not set up as government owned from the get go since they where de facto guaranteed by the government. The SBA and their loan program is not run as a GSE and borrowers pay a significant fee to fund the 80% (never 100%) guarantees.

    “4. Big bank purchases of MBSs.”

    Maybe I’m missing something, but I don’t see how you can say that derivatives are NOT a main problem then cite the purchase of derivatives by large banks as a major problem.

  27. Gravatar of Silas Barta Silas Barta
    8. November 2011 at 08:13

    @Scott_Sumner (and in regard to Paul_Andrew’s accurate points on the matter):

    Oh gosh, I am SO SORRY my brief sentence was vague enough that you could interpret it to refer to velocity. Let’s try it again:

    “…no one actually cares how many instances there are of a dollar being transferred to another person for a purchase in a given time period!

    Or how about:

    “…no one actually cares how many times dollars are transferred to another person for a purchase in a given time period!”

    I was referring to total dollar transfers, not average transfers over each dollar, which would mean I correctly refered to NGDP, not velocity, and making Scott_Sumner and W._Peden non-resonsive (and, if I may say, a bit whiny).

    Either way, it doesn’t change the point: no one cares about this number except to the extent they can infer something about stuff that does matter.

    Watch this: “You own a business selling nose cleaners. You learn that the total number of times that some dollar changed hands this year was 50 billion. Last year it was 49 billion. Should you expand or liquidate your business?”

  28. Gravatar of Silas Barta Silas Barta
    8. November 2011 at 08:15

    @Paul_Andrews:

    When I use Silas’ phrase “The number of times a dollar is transferred to another person for a purchase in a given time period”, I mean (and I presume Silas means) “The number of times anyone transfers any dollar to another person for a purchase in a given time period” – i.e. the total number of dollars spent for purchases in the economy.

    Correct. I’m glad someone got that right rather than uncharitably reading my statement the wrong way so as to dodge it.

    Know anyone like that, Scott_Sumner? How about W._Peden?

  29. Gravatar of W. Peden W. Peden
    8. November 2011 at 08:50

    Silas Barta,

    So I would say I understand your point, but I assume I haven’t since it is presumably argumentative and yet it doesn’t clearly contradict anything that anyone has said.

    Let’s try some more interpretation (not too uncharitable, I hope): imagine an economy where almost no economic activity takes place, except some minimal barter and two government bureaucrats exchanging buckets of freshly fallen rainwater at $500 a time at hundreds/thousands/millions/whatevers times per day. The statistics show that NGDP is growing at 5% and the central bank is pleased. Everyone else thinks that the economy sucks.

    As I understand it, Silas Barta, your point is that the fact that the central bank is facilitating the money supply such that NGDP growth = 5% in this economy doesn’t mean that this economy doesn’t have extreme problems. No-one cares that these two bureaucrats are trading these buckets of water. Talking about “the effects of expected NGDP growth” in such an economy would be silly.

    Is that an accurate interpretation of your point?

    (I’m not fond of talking about expected NGDP growth at all, since most people don’t have such expectations. There are expectations of future monetary policy & asset prices, which imply NGDP expectations.)

  30. Gravatar of Morgan Warstler Morgan Warstler
    8. November 2011 at 09:43

    Yes, Silas is technically correct, except…

    We do have a history of V, and surely it make little sense to talk about an example which blows it far far past anything that history has seen.

    To me it comes down to this, we could have a level target 3% (or even 4%) NGDP starting today, and for the next 20 years we’d experience far less inflation than we did over the last 20 years.

    That’s purely based on historical growth, but the odds of that growth continuing under such a plan, are to me, almost 100% guaranteed.

  31. Gravatar of Silas Barta Silas Barta
    8. November 2011 at 11:25

    Yes, W._Peden, that is an accurate representation of my point, although the situation does not need to be nearly so contrived to get such a divergence between NGDP and “the economic goodness we want and care about”. Any time you target a metric because of its past correlation with goodness, you are likely to see the correlation break down. Goodhart’s Law/Lucas critique and whatnot.

  32. Gravatar of W. Peden W. Peden
    8. November 2011 at 11:39

    So I want to separate some points-

    1. NGDP and “economic goodness” (I borrow your phrase; it’s probably the best for what we both mean here) can certainly diverge. However, I don’t expect any monetary system to guarantee economic goodness.

    1.1 There is one thing that worries me about all NGDP targeting, though, which is the relationship between the money supply and assets. Assets are not included in NGDP, which is only a measurement of final goods; I have found instances where NGDP was stable and asset prices were very volatile. This is not a specific problem for NGDP targeting, but it may be an inevitable product of any fractional-reserve banking system and must be weighted against the advantages.

    2. I think that Goodhart’s Law applies where there are agents with the incentive and the power to distort figures. With NGDP, I don’t think that central bankers have the power (though they have the incentive) and everyone as a whole has the power but not the incentive to go to the effort of collectively distorting NGDP figures.

    3. Since arguments for NGDP targeting (particularly Scott Sumner’s) tend to emphasise the effects of introducing NGDP targeting on expectations, I don’t think a straightforward Lucas critique is applicable. For example, I don’t think anyone can say that Scott Sumner’s NGDP futures market or George Selgin’s productivity norm are policy-invariant theories. Quite the contrary: they only work if people have forward-looking expectations.

  33. Gravatar of Silas Barta Silas Barta
    8. November 2011 at 11:51

    @W._Peden:

    1) Scott_Sumner claims that NGDP targeting can remove much of the economic badness that we currently have — that somehow, if a higher number of dollar-purchases were happening, then more economic wants would necessarily be satisfied on net.

    2) They most certainly do have the power: wait till the end of the God-decreed relevant period for measurement, and then print money and buy enough finished goods to top off the NGDP target. (Note the problem of periods: why much NGDP reach a per-year target? Why not a per day or per hour target? Per second? Oh, right, because sometimes holding money for the future is a *good* thing…)

    The Goodhart problem more generally is that if A has no causal power on B (despite and observec correlation, like between economic goodness and NGDP), then causing A will not cause B, but rather, will cause whatever extreme amounts of A cause. (These days it takes the form of, “Inflaion happens in good economies. So if we cause some more inflation now, our economy would be non-bad.” I can produce actual quotes like this.)

    And that’s the error in a lot of the NGDP targeting advocacy.

  34. Gravatar of W. Peden W. Peden
    8. November 2011 at 12:24

    1. Necessarily or presumably make things better? I don’t think that anyone denies that we can theorise situations where more NGDP doesn’t result in better conditions. At most, that tells us the rather unsurprising fact that NGDP targeting is not a panacea or (to paraphrase Keith Joseph) “Market Monetarism is not enough.”

    There are many problems outside of the disequilibrium between the supply of money and the demand for money; however, I see no reason to think that solving this problem would make the other problems worse.

    2. Can the Fed buy final goods? Also, this is not a case of Goodhart’s Law, because (a) it doesn’t affect the information content of the data, hence & (b) it is statistically detectable. If the deception is statistically detectable, then the incentive for the Fed to do it (assuming for the sake of argument that they could) is removed since they can’t deceive in order to cover their tracks.

    3. Regardless of whether or not Goodhart’s problem is actually the difference between correlation & causation (I think that Goodhart cases are possible even with genuine causal relations) you make a good point with which I (at least) partly agree. Let’s see if we can keep agreeing:

    (A) Inflation- defined as permanent increases in the price level- is not the mark of a healthy economy.

    (B) Inflation is not the cause of real growth. Arguments for NGDP growth based on the proposition that inflation is a necessary condition for growth are all wrong.

    (C) Short-run “inflation” in response to a supply-shock, if this is a consequence of producers raising their prices rather than an increase in NGDP growth above the trend rate of total factor productivity, is benign for the same reason that deflation caused by increased overall production is benign.

    (D) NGDP growth is a necessary but not a sufficient condition for growing national wealth.

    (E) An NGDP growth target like the one described in George Selgin’s “Less Than Zero”, with a trend of deflation & the denationalisation of money, would be preferrable to the currently predominant inflation-targeting systems.

  35. Gravatar of Silas Barta Silas Barta
    8. November 2011 at 13:11

    @W._Peden:

    1) Either way, I dispute the claim that it will be contributory toward making things better, for the reason that “buying is not good; good buying is good”. That is, purchases people make because they genuinely want the stuff are good, and indicate some Pareto improvement. You can no longer say this once you significantly manipulate people’s incentives (as NGDP targeting necessarily does) to the point where that purchase *becomes* good when it wasn’t otherwise.

    As an example of the dynamic (but not at necessary precondition for the effect I’m describing), if you have a hyperinflationary situation, people try to spend money the moment the earn it, but this is a sort of frantic “buy before your money withers up” effect, and should not be equated with the value of purchases of people *not* under such a desperation constraint.

    2) If “all options are on the table” to target NGDP, then yes! The thing is, the distortions don’t need to come from such direct purchases; any attempt to induce people to buy more than they otherwise would (see 1) runs the risk of bringing in “hollow” purchases that do nothing for the economy

    3) I agree that Goodhard effects can exist even when there is a causal relationship. For example, if A causes B, but A does not cause B when it is deliberately targeted by some new outside force. (Simple case in point: normally, possibility of having to pay damages causes me to buy insurance. However, if some third party devotes itself to making sure my out-of-pocket expenses increase right in line with my insurance coverage — perhaps in an attempt to fight underpayment of tort judgments — then the possibility of having to pay damages no longer causes me to buy insurance.)

    I agree with A-C but not D and E. E because of the general effect that such targeting introduces (see previous paragraph), and D because an economy can better satisfy people’s desires even if dollars change hands less often. (Extreme case: people build robot servants that harvest and give them everything for free.)

    I’m also interested in your thoughts on how to choose the optimal period for NGDP targeting, i.e., why have a per-second NGDP target, other than measurement costs?

  36. Gravatar of Jeff Jeff
    8. November 2011 at 13:18

    About the altruisism of unrelated and uncompensated kidney donors:

    To really encourage these donations, each donor could be given a guarantee that, should his remaining kidney start to go bad, he would be first in line to receive a transplant. Very few would ever need to use the guarantee, but it would likely increase the number of kidneys donated by a lot. Everybody wins.

  37. Gravatar of W. Peden W. Peden
    8. November 2011 at 13:41

    Silas Barta,

    (1) If you want to argue that ALL policies that affects incentives are not good, then fine. As it happens, I think that our desired balances of the medium of exchange are infinite, since it can be exchanged for an infinite number of existing & future goods.

    (2) I don’t think that all options should be on the table to target NGDP. Also, I think that monetary policy can only ever accomodate spending by the public; that is to say that it accomodates the infinite desire for the medium of exchange.

    (3) Do you have any actual cases of counterexamples to (D)? Also, why should the presence of an NGDP targeting authority remove the requirement for money to (in Hume’s phrase) oil the wheels of trade, which is Selgin’s intention*?

    In the hypothetical robot case, the robot-produced goods drop out of economic statistics (like charitable services) and presumably the trend rate of total factor productivity falls since leisure just became a lot more attractive. Keeping things mechanical, Selgin proposes a computerised approach to deciding the trend NGDP rate based on an assessment of the future trend of TFP. So the NGDP target is lowered in response, preventing the emergence of an inflationary trend.

    In answer to your question: I think that the time period is less important than attention to the trend. If NGDP overshoots one year, monetary policy should be tightened the next. Obviously some periods are going to be too short to be practical (e.g. 3 months) and some periods are too long (4 years or more and entire business cycles will regularly exist within the target period). Perhaps 1 year time periods have advantages over, say, 18 month periods: people are used to thinking with them and they correspond with tax years so the effects of fiscal policy can be easily estimated.

    I really do recommend “Less Than Zero”. It was the paper that won me over to deflationism.

    * Of course, a productivity norm might require less oil for the wheels in certain circumstances e.g. during endogenous booms when the demand to hold money falls.

  38. Gravatar of Silas Barta Silas Barta
    8. November 2011 at 14:07

    @W._Peden:

    1) I agree, which would seem to make the idea of using monetary policy to help people reach their optimal balances pretty pointless … no amount of cash will be “enough” for anyone.

    2) Okay, but Scott_Sumner does think all options should be on the table, or at least significantly more than there are now. But like in 1), how can you possibly accomodate an infinite desire?

    3) You said it’s a necessary condition, so I only need to show how it’s theoretically possible, which I did. That said, there should be some late 19th century economy with economic growth coinciding with sufficient deflation to have negative NGDP.

    I agree that a method of adjusting NGDP to account for TFP and, er, non-monetary provision of goods would be an improvement and significantly more robust, so I don’t have a criticism at the moment of NGDP targeting with this modified metric. However, a) that’s not NGDP anymore, which is inherently monetary, and b) my criticism is of Scott_Sumner, who does not advocate such an adjustment.

    Your handling of the time period issue is likewise more intelligent than any handling of the matter I’ve seen from Scott_Sumner, but it still seems to assume away the benefit of holding onto cash for longer than the period the CB uses, and will hurt people who want to (rationally) do so, and the economy in general to the extent that the desire to hold cash is motivated by real factors (e.g. uncertainty).

    And just to clarify, I don’t dislike deflation, I just dislike targeting some monetary metric on the basis of its historical correlation with economic goodness.

  39. Gravatar of Silas Barta Silas Barta
    8. November 2011 at 14:08

    Also, in regard to your 3): you’re right that money is necessary to grease the wheels of commerce, but from that it doesn’t follow that a greater amount of good commerce requires and increase in the amount of money.

  40. Gravatar of W. Peden W. Peden
    8. November 2011 at 14:55

    (1) Yep, which means that- given the monopoly of the monetary base still held by the central bank- a deflationist NGDP targeting regime, like a productivity norm, is the least bad of all words.

    (2) It can be accomodated to a degree. Also, “more than now” is different from “all”.

    (3) If there was, Rothbardians would have found it by now. There was some excitement about the US economy circa 1838-1842-ish, but it turns out that overall real GDP growth fell alongside NGDP growth during that period and any significant divergence between the two was due to extremely rapid immigration, which boosted national RGDP but not per capita RGDP.

    As for Scott Sumner, I think one day I will try to persuade him of the superiority of Selgin’s productivity norm.

    There is definitely a problem with time preference conflicting with central bank behaviour, which is probably why there is a relationship between good monetary proposals and the degree to which the central bank is disabled, which completely free market banking probably at the top of the heap. IIRC, Friedman came up with a solution to this problem that was even more deflationist than the productivity norm and did quite a bit of work on trying to find a way to get the long-run interest rate to be neutral.

    I suspect that anything short of free banking (where interest rates would be strictly determined by the intersection of the market supply & demand for credit) isn’t going to entirely address this problem in a satisfactory way. Like I said, NGDP targeting doesn’t solve every problem, including the existence of a monopolist controlling the monetary base and indirectly manipulating the supply of credit.

    I agree that a correlation with goodness should never be confused with goodness as such. Asset prices are correlated with periods of strong productivity growth, but I wouldn’t want that correlation to be the basis of monetary policy.

    I also agree that increases in the supply of money aren’t always needed for good commerce. Like I said, sometimes wheels don’t need oil at all; sometimes they may even need to have their wheels dried off. For example, the contraction of broad money in the mid-1990s was a good thing, since it poured water on what would otherwise have been yet another major boom & bust cycle and was instrumental in the length of the “Great Moderation”.

  41. Gravatar of Silas Barta Silas Barta
    8. November 2011 at 15:24

    Thanks for the replies, W._Peden. It seems we agree a lot more with each other (and you a lot less with Scott_Sumner) than I previously thought. All I want to add at this point is that holding on to cash is more of a liquidity/uncertainty issue (i.e. the less likely you will be able to find a comparative advantage, the more you will want to hold cash and other liquid/general use goods) than a time preference one, although high time preference likewise can make you blow all your cash (like a good Central Bank pawn).

    Your insights would be welcome on my blog, stop by if you get a chance.

  42. Gravatar of Peter N Peter N
    8. November 2011 at 17:58

    @ ssumner
    I agree that subprime mortgage system was an accident waiting to happen, and that Fannie and Freddy were partly responsible. How responsible depends on whether you like stupid and reckless lobbyists, or crooked, and clever moral hazardists. (Definition – One who seeks gain with no concern for risk, safe in the rational expectation that any losses will be borne by others).

    Fannie was bitten by The dual nature of the GSE’s. They had unhappy stockholders and were losing market share to the financial wizards of leverage.

    The investment banks needed AAA securities, and there weren’t enough. MBSs solved that problem (with the help of compliant rating agencies and a blind eye to the details of origination). Subprime mortgages were more profitable, and the banks believed they could hedge or sell off the risk. So loan quality was just an obstacle. This is a survey article from the St. Louis Fed. It’s a must read, if you want facts instead of spin.

    http://research.stlouisfed.org/publications/review/06/01/ChomPennCross.pdf

    Fannie started losing market share to banks willing to write higher risk loans. They decided they needed a piece of the market – $1.1 trillion in total. They did business with Countrywide. The subprime loans produced 60% of the GSE’s losses. The 2005, 2006 loans are radioactive.

    The banks were worse in every respect (worse default rate, lower standards…) , but none was near as big as Fannie. Still, without derivatives this would have been like the S&L fiasco, not the Great Recession.

    The banks found ways to get around their reserve requirements with off balance sheet vehicles and a way to inflate their money supply by increasing its velocity. The repo market did that. They made $1 do the work of $20. Then the smarter ones (like Goldman) successfully hedged their risks. Lehman and Bear got caught short.

    The deterioration of the subprime loans raised the repo haircut. Now banks needed to supply additional collateral or find money to repay their counterparties. The MBS market had tanked so they dumped their non-MBS securities such as AAA corporate bonds and disrupted the markets for those.

    This is the Gorton Q & A article that discusses the role of repos:

    http://www.zerohedge.com/sites/default/files/Q%20and%20A%20on%20the%20Crisis%20Feb%202010.pdf

    “The financial crisis was centered in several types of short-term debt (repo, asset-backed commercial paper, MMMFs shares) that were initially perceived as safe and “money-like”, but later found to be imperfectly collateralized. In this way, the crisis was a banking panic, structurally similar to centuries of panics safe, money-like instruments like bank notes and demand deposits”

    http://www.brookings.edu/~/media/Files/Programs/ES/BPEA/2010_fall_bpea_papers/2010fall_gorton.pdf

    Nobody knew who might be insolvent. It was right out of Lombard Street. The size of the original subprime mess wasn’t enormous, but 2007 was very different from 1987, when the damage was limited to the S&Ls.

    BTW add Greenspan to your list of causes. It takes a lot money to inflate this big a bubble. The $12 trillion in spare cash that was parked in repos had to come from somewhere.

  43. Gravatar of W. Peden W. Peden
    9. November 2011 at 03:08

    Silas,

    I’ll be sure to check your blog out. I agree that uncertainty is a big factor in the demand to hold cash.

  44. Gravatar of ssumner ssumner
    9. November 2011 at 13:42

    Paul, You said;

    “When I use Silas’ phrase “The number of times a dollar is transferred to another person for a purchase in a given time period”, I mean (and I presume Silas means) “The number of times anyone transfers any dollar to another person for a purchase in a given time period” – i.e. the total number of dollars spent for purchases in the economy.”

    No, purchases involving NGDP are less than 1% of all purchases.

    Rob, Interesting, but I don’t even consider 5% to be statistically significant, much less 10%.

    Eric, I don’t see where the derivatives made a big difference, but I was thinking of the more exotic derivatives. Haven’t MBSs been around for a long time?

    In any case the fundamental problem in this crisis seems the same as in the 1980s–smaller banks making lots of risky loans with FDIC-backed money. And the policy failure then, and in the subprime case as well, is that this sort of lending is legal, even with FDIC-backed deposits. Amazingly, Dodd-Frank did not address this issue.

    I agree that the quasi-public GSEs were a huge mistake, I’ve been arguing they should be abolished for two decades. The answer was not making them public, it was abolishing them.

    Silas, You said;

    “…no one actually cares how many instances there are of a dollar being transferred to another person for a purchase in a given time period!”

    Try again, that’s still not NGDP, it’s total transations, which is more than 100 times bigger than NGDP. If you are trying to say “NGDP” have you ever thought of just saying “NGDP?”

    You said;

    “Any time you target a metric because of its past correlation with goodness, you are likely to see the correlation break down. Goodhart’s Law/Lucas critique and whatnot.”

    Wrong. Goodhart’s law has nothing to do with “goodness” it’s about intermediate targeting. It has no implications for NGDP targeting.

    Jeff, It would help a little, but not much. Thousands of lives would be saved if we paid kidney donors.

    Peter, You said;

    “The banks were worse in every respect”

    If you mean the big banks, I totally disagree. The worst offenders were the small banks who required more than $100 billion taxpaper bailout. Ditto for the GSEs. At least the big banks paid back the TARP money.

    The recession was caused by the Fed, not the banking crisis.

    You said;

    “BTW add Greenspan to your list of causes. It takes a lot money to inflate this big a bubble. The $12 trillion in spare cash that was parked in repos had to come from somewhere.”

    No. You are confusing money and credit. The Fed creates money and the banks create credit. Money doesn’t go into markets, it goes through them. The only places it goes into is wallets and bank reserves.

  45. Gravatar of Paul Andrews Paul Andrews
    9. November 2011 at 14:26

    @Scott:

    You are correct, replace “purchases” with “final goods purchases”. The point Silas makes still stands despite these uncharitable misinterpretations of it (and he is not talking about the number of transactions, he is talking about the number of dollars).

    You did not respond to this:

    At http://www.nationalaffairs.com/publications/detail/re-targeting-the-fed you say: “This suggests that NGDP is useful not only as a predictor and indicator of trouble, but as a target for monetary policy.”

    Above you say: “Who ever said NGDP was used to measure economic health?”

    If it is not a measure of economic health then it cannot be an indicator of economic trouble.

  46. Gravatar of Peter N Peter N
    10. November 2011 at 08:19

    We seem to have different ideas about what money is. This is one of the words for which economists seem to have a number of different meanings, sometimes more than one per economist.

    Sometimes it means one or more of M0, MB, M1, M2, MZM and the late M3.

    And sometimes it means something which serves as a store of value, medium of exchange and unit of account.

    Thus Gary Gorton says:

    “Repo and checks are both forms of money. (This is not a controversial statement.) There have always been difficulties creating private money (like demand deposits) and this time around was no different.”

    This is how I use the term. Money in a context is anything commonly and effectively used as a medium of exchange in that context. AAA MBSs were money in the repo market, but not outside it.

    You’ve often said money was tight in some circumstance or other based on some set of economic indicators, but I have no idea how you were defining money.

    I’m also interested in your views on endogenous money. It seems clear that banks aren’t constrained by reserve requirements (c.f Bernanke and Blinder). So does high powered money exist? How is it defined? What is the secret of its power?

  47. Gravatar of Peter N Peter N
    10. November 2011 at 12:04

    I think this graph tells the Greenspan story.

    http://research.stlouisfed.org/fredgraph.png?g=3h7

    He steps on the gas to counteract the dotcom crash, but when he tries the brakes, they don’t work very well (M1 goes dead flat, M2 loses a bit of slope, mortgage debt rate of increase is unaffected). Wall Street just turns to the repo market, and the party continues.

    Refinancing homeowners took $1 trillion out of their mortgages. Every major financial institution wanted mortgages on which to practice financial alchemy. They weren’t prepared to take no for an answer.

    I’ve had to scale some of the curves to fit.

  48. Gravatar of ssumner ssumner
    10. November 2011 at 20:17

    Paul, You are mixing up levels and changes. I mean that a high level doesn’t indicate health in the way a high RGDP would. But very erratic NGDP creates problems, I’ve always said that.

    Peter, I define money as the base, and I define tight money as when expected NGDP growth is below target. In other words “tight” relative to the policy objective.

    Monetary policy played no role in the subprime bubble–it was bad regulation, moral hazard.

  49. Gravatar of Paul Andrews Paul Andrews
    13. November 2011 at 16:27

    @Scott,

    When you said NGDP is the actual “thing”, you didn’t qualify that with any references to levels vs. changes.

    The statement is a bombast.

    You agree that a high RGDP indicates health, and a high NGDP does not.

    “High” means “relatively large” in this context. In the context of GDP, the “relatively large” means “larger than the previous period”.

    So you would agree that an RGDP that is larger than the previous period indicates health, and an NGDP that is larger than the previous period does not necessarily indicate health.

    Therefore you would agree that a growing RGDP indicates health and a growing NGDP does not necessarily indicate health.

    Therefore the statement ‘NGDP is the actual “thing”‘ is, as I have said, empty rhetoric.

    It seems that your thesis is predicated on the concept of maintaining a level of NGDP growth regardless of the level of inflation (particularly when there are shocks to the economy). It seems you base this on the notion that historically when NGDP has been at 5%, inflation had been around 2% and there has been 3% RGDP growth. Have you considered that this history occurred during times of credit expansion and will not necessarily repeat in a deleveraging scenario? Have you considered that debt to GDP (private + public) cannot physically grow forever, but that it has been growing continuously over your historical time frame? Have you considered that the very act of maintaining NGDP growth at 5% in the past (instead of aiming for 0% inflation) may have lead to the over-leveraging and consequent deleveraging? Have you considered that in an NGDP futures market, the Fed may need to print unlimited amounts of money to maintain contract prices? At what point would you have the Fed default on its obligations to maintain the contract price? (Or would you let them just print to oblivion?)

    I would be interested to know when you would abandon an NGDP target? For example, if a large city, if 5% of the population of the US needed to be evacuated to another country, thereby reducing RGDP by 5%, to keep NGDP on its former growth path you would need an additional 5% inflation. It seems that you would recommend that approach. What if the population halved? At what point is your NGDP target abandoned?

    What do you see as the flaws and limitations of your own monetary policy recommendation?

  50. Gravatar of Scott Sumner Scott Sumner
    13. November 2011 at 19:45

    Paul, I’ll take the easy question first. I actually favor targeting per capita NGDP, but never talk about it to save time. In any case the US poulation growth is very steady. But that answers your city evacuated example.

    The harder question is RGDP vs. NGDP. I understand why you see a conflict. Let me use an urban analogy. I think New York City is the actual thing, and the New York metro area is a arbitrary social construct, like the CPI or RGDP. The suburbs of NYC gradually become more dispersed as you move farther from the city. Who’s to say exactly where the metro area ends? In contrast, NYC is very precise–5 boroughs with well demarcated boundaries.

    Yet even though the metro area is more arbitrary and imprecise, I think the metro population is a better measure of the size of a city than the population within city limits. For instance, Boston has a relatively small population within the city.

    By analogy, RGDP growth is much fuzzier than NGDP growth, but with all its faults is a better estimate for growth in living standards, especially for cases like Zimbabwe.

  51. Gravatar of Paul Andrews Paul Andrews
    13. November 2011 at 23:43

    Scott, so if the population reduced by 5% you would favour targeting 0% absolute NGDP growth in that year (i.e. maintain 5% per-capita NGDP growth).

    What if RGDP drops by 5% for some other reason – e.g. debilitating chronic disease? What if it drops by 20% for some real world reason other than population decline? Would you maintain the 5% NGDP growth target (meaning 25% inflation in this case)? Where do you draw the line?

  52. Gravatar of Scott Sumner Scott Sumner
    14. November 2011 at 18:13

    Paul, I’ve stated many times that the optimal policy is a nominal wage rule, but that’s not politically feasible. So you’d look at wages in each case.

  53. Gravatar of Paul Andrews Paul Andrews
    15. November 2011 at 01:36

    Scott,

    How would that work? If it’s politically infeasible to have a nominal wage rule, what would make the Fed look at wages once they are fully committed to the NGDP futures market?

  54. Gravatar of ssumner ssumner
    15. November 2011 at 13:16

    Paul, I meant you’d consider wages whenc designing the optimal NGDP regime. Thus the argument for per capita NGDP is that it better stablizes wages. You could go even further—NGDP per working age adult. That gets you even closer to wages. Hourly wages are best, but probably not politically feasible.

    But you are right, you don’t want to change the regime once it’s in place.

  55. Gravatar of Paul Andrews Paul Andrews
    16. November 2011 at 05:20

    Starts to sound much closer to wage inflation targeting than NGDP targeting.

  56. Gravatar of Scott Sumner Scott Sumner
    16. November 2011 at 16:53

    Paul, Years ago I published a paper arguing wage targeting is optimal. But it doesn’t seem politically realistic, so I favor NGDP targeting as second best.

  57. Gravatar of Paul Andrews Paul Andrews
    16. November 2011 at 22:08

    Scott,

    You’re basically defending your view of the world rather than NGDP per se.

    The issue is not whether Scott Sumner is correct, the issue is whether NGDP targeting would have positive or negative effects on the economy.

    The fact that you prefer wage inflation targeting does not assist the case for NGDP targeting. Whether you think NGDP is second best has no bearing.

    You agree that there are flaws in NGDP targeting. I don’t see these flaws mentioned when you are advocating NGDP targeting, or when you are reporting on others’ advocacy for it.

  58. Gravatar of Scott Sumner Scott Sumner
    17. November 2011 at 19:15

    Paul, I’ve made all these points numerous times in my blog. Blog posts are short–you don’t do an entire dissertation on all aspects of NGDP every time you do a blog post.

  59. Gravatar of Paul Andrews Paul Andrews
    17. November 2011 at 20:55

    No I agree. Conversely catch-phrases like “NGDP is the ‘actual thing’” are disingenuous if you believe NGDP targeting is sub-optimal.

  60. Gravatar of Scott Sumner Scott Sumner
    18. November 2011 at 18:56

    Paul, No, when I say it’s the actual thing, that has no bearing on whether it should be targeted. The price of gold is also an “actual thing” but I don’t favor targeting gold

  61. Gravatar of Paul Andrews Paul Andrews
    19. November 2011 at 00:42

    It seems what you really meant then was “NGDP is *an* ‘actual thing’”.

  62. Gravatar of Scott Sumner Scott Sumner
    19. November 2011 at 18:38

    Paul, The universe contains many “actual things”. I meant compared to inflation and RGDP.

Leave a Reply