10 Nov 2011

A Minor Quibble With the Whole NGDP Discussion

Economics 83 Comments

I generally like what Noahpinion says on the GDP debate, but even he is conceding too much to Sumner et al. Before I was letting it go, because I figured it was a harmless simplification, but thus far I’ve only seen one random guy in the comments mention the problem. (Firefox ate my tabs when it upgraded, so I lost the link to that guy.)

Anyway, here is Matt Yglesias telling us why NGDP is “the real thing” as opposed to that artificial construction of BLS economists, “inflation”:

Something worth flagging for the NGDP conversation is that because both real GDP and inflation are familiar subjects of discourse, it seems natural to many people to take them as “given” and understanding NGDP (= GDP + inflation) as somehow constructed and exotic. But actually NGDP is, relatively speaking, the simple quantity here. It measures total spending in the economy. You count everything, add it all up, and you’ve got your NGDP. It’s actually just like the GDP you may have learned about in your intro textbook: Household consumption plus exports plus government purchases plus investment minus imports. Social construction enters the picture when we try to move from this nominal quantity to the allegedly “real” one.

No, Mr. Yglesias, notwithstanding your clever title reference to The Matrix and/or a book I never read, you are simply mistaken. NGDP is not “total spending” in the economy, and it is wrong when people write that it is equal to MxV, if V truly means “the average number of times a dollar changes hands in the economy during the time period in question.”

For example, if I spend $1000 buying shares of stock from somebody, that is “spending” and those $1000 change hands. But it doesn’t raise NGDP by $1000.

Even more serious, when the farmer sells wheat to the miller, who in turn sells flour to the baker, who in turn sells whole loaves of bread to the retailer, etc. etc., NGDP doesn’t capture the “total spending” in that stream of transactions. No, it only counts the “value added” at each stage.

So there are two things here:

(A) At the very least, there is a measurement issue. It is actually a lot more complicated to calculate “nominal GDP” than to simply add up “total spending.” You have to be a trained economist to even know how to begin doing it.

(B) There are actually conceptual issues too. For example, it is my understanding that if the baker spends $1 million on flour that he uses up that period making bread, that doesn’t count in NGDP. However, if he spends $1 million on buying a new oven that will last for ten years, then it does count in NGDP as “investment.” So, what happens if we switch our time period from one year to ten years? Does NGDP over the ten-year span suddenly drop by $1 million, because the oven went from fixed capital to circulating capital? (I have never gotten a good answer to this one, because everybody I asked said, “I don’t know Bob, I hate GDP accounting.”)

Last thing: For those of you think “this is the best Bob’s got,” I am working out some of these tangential issues as I prepare my most devastating assaults on Scott Sumner’s worldview. I have not yet begun to argue.

83 Responses to “A Minor Quibble With the Whole NGDP Discussion”

  1. Daniel Kuehn says:

    Why doesn’t the flour count in NGDP exactly? The value of the flour is part of the value of final goods and services, right? Investment is counted in the year of purchase, but why would you measure it otherwise? I have lots of durable goods that I continue to get value from, but “getting value from X”, and “expending funds on X” are two very different things – so it seems right that we measure investment in the year of purchase. Housing works a little differently – not sure if that’s good or bad.

    As for this: “For example, if I spend $1000 buying shares of stock from somebody, that is “spending” and those $1000 change hands. But it doesn’t raise NGDP by $1000.”

    That $1,000 is counted as income to you from an income approach, right (you buy it out of labor income and labor income is counted, right?)? I don’t know what the analogous component would be in an expenditure approach, though (and it would seem like there would have to be). Anyway – I hate GDP accounting.

    • Daniel Kuehn says:

      I’m not disagreeing with you – interpreting the “Y” in the quantity theory as GDP is problematic – just trying to think through some of the finer points and rescue GDP as a useful concept.

    • Bob Murphy says:

      DK, you might count the capital gain on a stock sale as part of GDP that year; I’m not sure. But clearly they don’t count every sale as part of GDP. If I’m reading this table correctly, just on the NYSE there are some $50bn in trades per day in early 2009. So that’s almost the size of the whole US economy right there, just on the New York Stock Exchange.

      • Daniel Kuehn says:

        Right – sorry if I was confusing, and that’s why I followed up. The quantity theory is a reference to transactions, and I’m agreeing just subbing in “GDP” isn’t really correct.

        Stock purchases are captured in GDP insofar as income expended on those stocks is counted. But if we are measuring gross income or gross product (i.e. – GDP) we don’t want to measure all $50 bn in stock transactions” because that would be double-counting the same dollar of income/production.

        Maybe a better way for me to put it is this: income spent buying stocks is counted (although I don’t know where it enters into the so called “expenditure approach” to GDP accounting). But it is not the case that every single stock market transaction is counted.

        After earning $100 from my employer I could write a $100 check to my own checking account in two different banks and transfer the same $100 to a different bank every day for the whole year. That $100 in gross income is counted in our gross income statistics. The $36,500 in gross transactions are not counted in GDP because GDP is not GDT – it is GDP and always has been.

        • Bob Murphy says:

          Oops DK I hadn’t seen this comment when I wrote my other response. So maybe we are closer in views than I realized. Anyway, sorry if I’m lecturing you with stuff you knew in preschool.

          Anyway, this all underscores my point that Yglesias and Sumner are wrong if they think it’s a trivial addition problem to calculate NGDP, whereas it takes controversial assumptions and expert economists to calculate (price) inflation.

          • Daniel Kuehn says:

            My view is it is a non-trivial addition problem to get NGDP (and it is an important number). It requires a non-trivial controversial assumptions to get inflation (and it is an important number), and it is a trivial multiplication problem to get real GDP from both of those (and it is an important number).

      • Daniel Kuehn says:

        So… with the caveat that the income spent on stocks is in GDP, I agree with you on that.

        I think I disagree with you on flour and investment purchases.

        • Bob Murphy says:

          Whoa hang on a second. It’s not the income spent on stocks that goes into GDP; it’s the income you earn from selling stocks at a capital gain that (theoretically) should go into GDP for some period.

          If you earn $100,000 from your employer, and then spend it on stocks, that’s $100,000 in GDP all right, but it’s not because of your “investment.” It’s because of your labor.

          As far as the flour, that’s standard stuff. Samuelson in his classic textbook goes over this kind of example, saying you have to watch out for double counting. E.g. if the farmer sells the wheat for $10, then the miller sells the flour for $50, then the baker sells the wholesale bread for $75, and the retailer sells the bread to the consumer for $100, then GDP from all of this is just $100. You can either count the finished good of the bread, or you can add up how much value is added at each stage (e.g. the miller buys for $10 and sells for $50, so he added $40 in value at that stage). But for sure you don’t say GDP is the sum of total spending, i.e. $10+$50+$75+$100 = $235.

          (Incidentally, this is one of the standard Austrian criticisms of GDP accounting. It’s why we have such absurdities as saying “the US economy is driven by 70% consumption” when no it isn’t, if we are counting total spending.)

          • MamMoTh says:

            What part of spending on new goods and services you didn’t understand?

            When you sell your old car it doesn’t count as GDP either even if you make some gain.

            • Daniel Kuehn says:

              He understands this perfectly, MaMoTh. And I still think he and I are on the same page – I may just be being unclear.

              Total transactions are important – that helps us understand the quantity theory and the role of the money supply.

              Total production/income is also important – we measure this with GDP.

              You don’t get GDP by adding up the transactions for intermediate goods. YOU DO include all intermediate goods in GDP (just not their full transacted value).

              • marris says:

                > Total transactions are important – that helps us understand the quantity theory and the role of the money supply.

                Be advised, some MMTers _don’t_ understand the quantity theory. Or they “understand and reject” it? Whatever.

              • MamMoTh says:

                I am sure he does understand it.

                Then he has an issue with flour not being accounted for, later he shows it is accounted but not twice?

                Stocks are neither goods nor services.

                GDP is quite trivial to define, but complex to compute.

          • Daniel Kuehn says:

            I earn income – I spend a certain amount of income on stocks or any other security. That INCOME is in GDP.

            The nominal value of all of the transactions I may make is not in GDP, because GDP is gross domestic PRODUCT, not gross domestic TRANSACTIONS. One income can make multiple transactions.

            re: “But for sure you don’t say GDP is the sum of total spending, i.e. $10+$50+$75+$100 = $235.”

            Exactly. That’s why I’m trying to differentiate between gross domestic product and gross domestic transactions.

            The quantity theory works of off transactions. As far as I know, we don’t really have good measure of that.

            • Silas Barta says:

              I earn income

              Yes, and it’s a sad indictment of our current society that one can earn income from what you do.

              Of course, the same could be said for me, to some extent.

    • MamMoTh says:

      Buying outstanding shares or cars do not count toward GDP.

      Not sure about newly issued shares.

  2. MamMoTh says:

    My understanding is that if the baker sells his bread (a final product) then the $1 million is counted in the price of the bread he sold. And if he doesn’t use all the flour it counts as investment in inventory.

    And M*V is useless anyway.

    • Bob Murphy says:

      Mammoth, what if the government stopped buying bread? How could the private sector produce net calories?

      • MamMoTh says:

        The government is not the monopolist producer of net calories. I thought you knew at least that.

        • David S. says:

          You’re obviously very wrong to make assumptions about what Bob knows, since he knows almost nothing.

          • Joseph Fetz says:

            Quite a statement from the “one” that has graced us with his uncompromising foresight and knowledge by his own self-attribution.

            You always seem to talk a lot of $hit, but you never seem to come through with any ideas worthy of value.

            It would be nice if you could further elaborate on WHY you believe that Murphy is unknowledgeable in the field of economics, but you never seem to do that. Instead, you state your empty rhetoric that is heavily devoid of knowledge and theory.

            How about this, at some point in your correspondence, you actually give us (or, anybody for that matter) reason to take your words seriously. Until then, these will be my last words to you… you will be ignored otherwise.

  3. Brent says:

    “NGDP is not “total spending” in the economy”

    Austrians have long pointed this out, particular economists like Skousen and Reisman. Given what GDP is trying to measure (the output of final goods and services — Income), it doesn’t even make sense that it would calculate anything close to total spending in the economy.

    • Daniel Kuehn says:

      Right – it’s spending on final goods and services. Isn’t that implicit in the definition?

      I think most people have a pretty good grasp of what GDP is and is not if they’ve taken any econ. That’s drilled in in intro macro (perhaps with some lingering confusion on these esoteric issues of what the BEA does with certain things).

      I think what isn’t taught that well isn’t GDP – it’s the quantity theory. When people teach the quantity theory they often are vague or downright inaccurate regarding Q/Y/T. I’m sure I’ve fudged the quantity theory before.

      The logic of the quantity theory still works OK if you sub in GDP for Q/Y/T – it just doesn’t hold as an accounting identity.

      • Brent says:

        C’mon, Daniel. You listen to the news. Economists and business journalists are always saying “spending” and equating it with GDP. Increase spending, increase GDP.

        And I know I pound it in the classroom, but how many others do? I doubt any of my former professors have ever mentioned it in a classroom. They certainly didn’t when I was around.

  4. marris says:

    Nice catch Bob.

    When NGDP targeting fails to bring about good results, I wonder if we’ll see people clamor for total spending targeting. After all, why should the Fed be forced to “stab in the dark” ?

    Silas, have you assembled a Do Not Target These Economic Statistics list?

  5. bill woolsey says:

    Buying stock isn’t counted as GDP.

    Trading newly issued stock isn’t counted as GDP.

    Capital gains on selling stock isn’t counted as GDP, (as part of income.)

    It is just spending on newly produced final goods and services. Income in GDP
    accounting is based on the flow concept, not the increment of wealth concept. So
    no capital gains (or losses.)

    However, the intermediate goods do get counted as changes in inventories.

    If you want, you could treat of all of them being counted when they are purchased.
    And if nothing further happened, they would count. But they get counted as a negative
    as they are used up.

    So, $10,000 worth of wheat is purchased. (GDP.) Oh, but $9,000 is used up in
    producing flour. So, this year, that is -$9000 “spending” on wheat. But there is
    $9000 spending on flour by the millers. If they sell $8000 of it, to bakeries, then
    that $8000 counts, but it is balanced by a -$8000 “spending” on flour by the millers.

    So, you plus and minus through the production chain, and in the end, there is some amount
    of sales of bread at the grocery stores, and no negative spending when it is eaten.

    If inventories of wheat or flour were reduced during the period, these count as negatives.
    But any incease in those inventories (purchased, but no negative when sold to the next stage,)

    I think the quantity relationship is MV = Py.

    I don’t use the transactions approach at all.

    V = 1/k and k is the demand to hold money as a fraction of real income. (or nominal income, if you like.) If you want to say it is how often a dollar is spent on final goods and services (including inventories of intermediate goods,) you can. But I don’t think that is all that useful.

    I am not sure about short lived capital goods and multi-year GDP calculations, but my first pass guess is that GDP is impacted, but not NNP. Now they are depreciated, but in the alternative measure, they wouldn’t be.

    GDP includes capital goods that are produced but just replace the capital goods that are wearing out. NNP only leaves net investment, taking out depreciation.

    However, I am pretty sure that the trend growth rate of real output would not be impacted by this shift. What is really happening doesn’t change at all, of course. The numbers might be different. But 3 percent growth in nominal GDP would still be consisent with stable prices even if capital goods with two year lives were accounted for as being used up in the 5 year GDP measurement. If you changed the measurement approach, on the other hand, and didn’t adjust the rule, then it would make a difference.

  6. AP Lerner says:

    This entire debate over targeting NGDP would be comical if it was not so serious and actually getting traction at the Fed. And the fact this is getting traction at the Fed is a sad statement for the economics profession. NGDP targeting (or mass QE, let’s be honest about this) is nothing more than a continuation of the neo-liberal nonsense policies that got us into this mess. The continued lack of basic understanding of the monetary system is depressing. The idea the Fed can target NGDP is about as comical as the Fed targeting the money supply. Mr. Murphy, with all due respect, you and Scott Sumner sound like two ancient intellects debating over how flat the earth is, with each of you claiming a continued flattening by the day. Meanwhile Pythagoras is sitting in the corner, correctly claiming in reality the earth is round, I mean, correctly claiming the public sector deficit is too small and not meeting the savings demands of an overtaxed private sector.

    • MamMoTh says:

      That is an issue with the Fed being able to target NGDP as the NGDP target advocates.

      But not with targeting NGDP which could be done at the fiscal level.

      • Major_Freedom says:

        What should the fiscal level people target? Aren’t they subject to the same limitations?

        • MamMoTh says:

          NGDP growth of 5% for example as Sumner wants?

          Fiscal policy is not limited like monetary policy, other than by self imposed constraints.

          Of course the Fed could start buying pencils and submarines and rebuild the infrastructure, but that amounts to fiscal policy, which should be in power of congress not the Fed.

          • Major_Freedom says:

            NGDP growth of 5% for example as Sumner wants?

            No, I mean what real things on the side of supply should the government target for spending. Not what rate of the growth in NGDP should be.

            Fiscal policy is not limited like monetary policy, other than by self imposed constraints.

            True, but it’s still limited by another sense, the sense that both fiscal and monetary policy are limited by, namely the laws of scarcity and of opportunity costs.

            Of course the Fed could start buying pencils and submarines and rebuild the infrastructure, but that amounts to fiscal policy, which should be in power of congress not the Fed.

            So wait, is NGDP targeting a monetary policy or is it fiscal policy? I thought it was a monetary policy. That’s what all the NGDP crowd is saying the Fed should be doing.

        • AP Lerner says:

          “What should the fiscal level people target?”

          Full employment for anyone willing and able to work

    • Major_Freedom says:

      Sooooooo…..free market in money then? No? OK, then what else can a monopoly issuer of money do to strike a balance between hyperinflation and economic growth with stable inflation? What should they target? How will they calculate it? Isn’t ANY monopoly issuer of fiat money subject to the same observation/calculation problems as your “not looking at empirical anything” Pythagoras sitting in the corner?

      There is a limit to what you can do with “public sector surplus is equal to private sector deficit”.

    • marris says:

      > NGDP targeting (or mass QE, let’s be honest about this) is nothing more than a continuation of the neo-liberal nonsense policies that got us into this mess.

      Ugh. Exactly what part of mass QE is _not_ MMT-friendly? Aren’t you really just objecting to the Fed buying assets from banks? If the CB bought [or rented] assets from non-banks, then this would be functionally equivalent to fiscal policy, right? So your objections are really just “political” objections.

      Yeah, I went there.

  7. kavram says:

    This actually explains alot regarding the tension between Austrian economics and using GDP as a barometer of economic health. Since GDP only measures the market value of FINAL goods and services, it must temporarily drop in order for the economy to truly grow (at least after a bubble). That is, every dollar that is used to buy raw inputs/materials/capital detracts from GDP, since they’re not considered “final” goods, whereas every dollar that is blown on consumption boosts GDP, but detracts from the capital stock that provides real economic growth.

    The Keynesians/mainstreamers overlook this necessary readjustment period, which I guess is why they focus so much on consumption and “spending”

  8. Rob says:

    Wikipedia seems to explain it quite well


    Seems an odd discussion to be having though – for me the real issue is whether or not having the CB try to stabilize AD is a good idea , not what the most accurate measure of AD is.

  9. RPLong says:

    What a coincidence to read that #2 point of yours this morning, Mr. Murphy! I was thinking almost the identical thing this morning while I was considering Sumner’s recent post on “savings.”

    I was thinking, on an aggregate level, is “saving” even possible? If I decide to save $1000, I put it in the bank. The bank lends it out or invests it otherwise, it doesn’t just lock my money in a room. So the only way I can *really* save a given quantity of money is to keep it literally under my mattress.

    The only sane way to view saving is as a mechanism for intertemporal substitution, i.e. delayed consumption.

    At that point, it becomes a question of periodicity. If I delay consumption for a year, and then consume all of my savings the following year, the 2-year average will be the same as if I spent it as fast as I got it. Only the 1-year averages change.

    Similarly, if you look at a hypothetical Ruritania consisting of 1 Francisco D’Anconia and 99 middle-class misers, Ruritania’s annual savings will perhaps be negative, because the millionaire spent more than he made, AND more than all the misers managed to save over the course of the same year.

    This is precisely why aggregate models don’t tell us anything substantial. It really comes down to arbitrary assumptions that make wild theoretical swings possible.

  10. Silas Barta says:

    What about the “private citizens fix NGDP” scenario?

    Let’s say Bob pays me $100 (cough cough) for “friendship services”. Then, I “buy” $100 of friendship services from him. Then, he buys $100 of friendship from me again. We repeat this back and forth until all these “purchases” sum to our desired NGDP target.

    Is each of those purchases counted toward NGDP? Under what conditions? Would we be taxed on the “profit” of each sale of friendship services? (Assume that the receipt of previous friendship was not necessary in order to sell friendship services.)

    • Major_Freedom says:

      I think you just hit on something that made me just think of something:

      GDP calculations are inexorably intertwined with taxation.

      Friendship services would generate practically infinite GDP if that $100 was not taxed. Only if it is taxed can it be applicable to GDP, because then the first $100 payment would be taxed (say 10%), meaning $10 would have to go to the government, leaving you with $90. Then when that $90 is paid the other way for friendship services, $9 is taxed, leaving $81. And so on. Eventually there will be no more friendship services capable of being sold.

      So there is a sort of multiplier here where an initial $100 is spent, and successively smaller amounts exist to be spent.

      It would be $100 + $90 + $81 + … + $0 = $1000.

      So GDP here is maxed at $1000 when taxation is 10%.

      But that tax money isn’t just hoarded by the government, it’s spent, and a total of $10 + $9 + $8.1 + … + $0 = $100 of taxes and hence spending is made.

      So total GDP is $1100.

      If friendship services are not taxed, then GDP could get as high as however many times you exchanged $100, which can be as high as you are capable of making an exchange of $100. If you had a computer, you could engage in high frequency trading like they now use on Wall Street, and over a period of a year, it would be 4 GHz (or however fast the computers can do this), times the number of seconds in a year.

      If we could trade something, anything, tax free, that is included in NGDP, we can stop inflation! Think about it. The more we add to GDP spending through repeated trades, the less the government can inflate. In fact, if we make enough repeated trades, and increase our share of NGDP enough, the government will have to deflate the money supply.

      I think you just blew up the viability of NGDP.

      • MamMoTh says:

        NGDP is a flow, money supply is a stock.

        Government can raise or lower taxes to target NGDP.

        What did you blow up?

        • Major_Freedom says:

          NGDP is a flow, money supply is a stock.

          Very good.

          Government can raise or lower taxes to target NGDP.

          Suppose lowering taxes didn’t succeed in raising NGDP, because the money went to uses that are not counted in NGDP.

          Suppose that the economy was such that the government would have to lower taxes to zero, or a negative rate, to target a desired NGDP rate of growth, thus necessitating inflation of the money supply.

          What did you blow up?

          I didn’t. Silas did. Silas even said that Sumner didn’t have an answer.

          • Silas Barta says:

            Crap, now I have to go dig up which exact exchange Scott_Sumner’s hasty dismissal was on…

      • Silas Barta says:

        I’ve actually brought up this exactly point to Scott_Sumner and gotten his response on his blog. He basically says, “but most transactions aren’t like that, I’m just going with normal stuff”. (i.e., “Ignore gaming of my plan.”)

        • Major_Freedom says:

          Ironically, that is the same mentality that guide’s the Fed’s targeting of consumer goods and ignoring capital goods and stocks, that Sumner et al are allegedly trying to overcome with NGDP.

  11. bill woolsey says:

    GDP figures include capital goods and intermediate goods.

    GDP doesn’t have to fall at least temporarily in order to grow.

    As I already explained, the production of intermediate goods is counted as inventory investment.

    When they are used up, it is negative inventory investment, but if that didn’t happen, then GDP doesn’t fall because resources were used to produce them.

    Generally, I think that arguments that GDP are an imperfect measure of well being tell us anything about the desirability of nominal GDP targeting. For example, whether or not the value of government output is properly measured is irrelevant. The point of nominal GDP targeting isn’t that there is an ideal level of nominal GDP. It is rather than keeping it on a stable growth path is desirable.

    There is no need for I pay you to be my friend and you pay me. It is simpler. I pay you $1 trillion to pick up a piece of trash. It is a credit transaction, and you owe me in one year. When one year comes up, I refuse to pay. (Ha, ha, it was a joke.) Nominal GDP when up by $1 trillion. A trash pick up service valued by the buyer at $1 trillion.

    As a practical matter, why would you believe that the BEA economists would count this? Why would passing back and forth a $100 bill make it more likely to be counted? To me, the passing back and forth the money for imginary goods is presumably based upon equation of exchange thinking. Running up the velocity.

    Of course, it would be much easier to fool the BEA guys if you were actually in business to produce things. So, you just get your ordinary buyer to agree to a higher price, payable later, with secret side deal that the premium will be forgiven when the bill comes due. This gets reported with all the other final sales data. It isn’t like someone is setting up a new firm and trying to get the BEA to recognize its output.

    And while some of this might have tax implications, not always if you don’t really collect.

    Of course, GDP is a really big number, and anyone trying to manipulate the numbers would have to do alot.

    Also, we already have index futures markets. Most of them are for prices, and so (like in my example above) fake transactions at different prices would impact the index and so generate profits or losses on the futures. I suppose with NGDP, you could create phantom quantiies as well as prices. So, I report that I sent 8,000 bolts of cloth to my customer. And later, he just says he didn’t get it, and so there is no payment.

    • Silas Barta says:

      I don’t think you understand the “friendship swaps” agreement. None of it has to involve credit or loan forgiveness, or failure to deliver. Everything looks from the outside like a legit purchase. The BEA has no reason to exclude such transactions, as long as I report them, and yet they are obviously not the kind of nominal spending that “market monetarists” want.

      Once you factor in the NGDP futures markets, as Bob_Murphy mentioned, people have a huge incentive to manipulate NGDP figures. (Your comparison to existing futures markets fails because what Scott_Sumner proposes is more like a prediction market than an actual purchase of output.)

      econometric + possibility to manipulate metric + incentive to manipulate metric = new deviation of metric from past “economic goodness” it was correlated with

      inability to appreciate the above phenomenon = poor economic understanding = misguided economic policies = Lucas/Goodhart critique

      (Btw, learn 2 reply to comments to directly.)

  12. bill woolsey says:

    Implicit in your argument is that because of value subjectivism, your friendship services are just as legitmate as any other service. You are implicitly assuming that you have a “right” to have he BEA count them.

    I think that they won’t. Why not try it? I am telling you that there is no need to pass money back and forth. Nothing about that legitimizes the transaction.

    It is often the case that goods and delivered for a price, the value of the good is counted but then the buyer never pays. It is because they go bankrupt and cannot. Now, if we do a fake transaction, then this is a scam.

    Well, you are simply wrong that I your friendship swap is anything but a scam. Why do it?

    Now, if people really liked doing this for some reason and continue to do it at the same rate, then it would have no impact on nominal GDP targeting. On the other hand, if there is a change in the level of friendship payment swapping, and there were nominal GDP targeting, then what would happen is that it would require a decrease in the level of prices and wages for other activity.

    The easiest approach would be to ignore the nominal “output” of this industry.

    And, of course, I could do that. Since there is a quid pro quo, in that I sell you friendship and you sell me friendship, buying your friendship is an input used to produce selling you friendship. (That is, I only get to sell you my frienship if I bought yours.) And so, my profit from the activity is zero.

    For example, suppose a retailer buys bread from the wholesaler. Then, the retailer returns it. Then he buys it again. Then he returns it. Then he buys it again. Is output increasing? No, it is the same output changing hands multiple times.

    Well, instead of producing more and more friendship, it looks to me that I am buying your friendship so you will buy mine. It is like buying bread from a wholeseller to sell it to a consumer.

    Are sevices different? Suppose we have a contract for a gym membership. It is transferable. The gym provides the services and people to to the gym. But they membership is sold 10 times. Does that expand the service by a factor of 10? No, it is just changing hands. Suppose just the two of us can’t make up our minds. We buy and sell it back and forth 10 times. Is there more gym services? No.

    Anyway, why not try it. Get a friend to play your frienship game, call the BEA, and see how to report the output of the game. Tell us what they say. Do they say, thank you. That is how GDP is measured. People call us up and tell us what they produced.

    • Silas Barta says:

      I think you’re over-complicating this, bill_woolsey. Let’s say Bob buys the friendship from me. His personal corporation records “$100 purchase of friendship”. My personal corporation’s records “$100 income from sale of services.”

      Later, we report the opposite.

      How does the BEA know we were gaming the system? All they know is that we kept buying services from each other.

      The easiest approach would be to ignore the nominal “output” of this industry.

      Not an option when people are deliberately doing this to win on prediction markets and the system relies on people not doing it. At the very least, it introduces the very subjective judgments into calculation of the metric that market monetarists are claiming they can avoid.

      I think you have a hard time understanding that once you predicate big decisions on certain numbers, people start to game those numbers.

      • Rob says:

        It seems that it is possible (though unlikely in practice) for fraudulent businesses to “game” the system for calculating NGDP and distort the results of an NGDP-targeting regime.

        It seems possible (and quite likely in practice) that the government might game its own NGDP-numbers for political ends.

        In a free market system where profit-maximizing fractional reserve banks stabilize AD without the need for any targets then this possibility of fraud goes away.

        • Silas Barta says:

          In a free market system where profit-maximizing fractional reserve banks stabilize AD without the need for any targets then this possibility of fraud goes away.

          No, it doesn’t, because:

          1) People can deceive any AD-targeter (public or private) the same way.

          2) People have similar incentives to induce runs on FRBs.

          3) No free market bank cares about optimizing AD, nor need its actions be therewith correlated.

          • Rob says:

            In essence FRBs are able to pick up the signals of increased demand for money (which correlates to falling AD) from their customers account balances and transactions and increase lending activity as a result. They do not need to care about AD, just their bottom-line and as there is not target there is no-one to deceive.

            Not sure about your 2) – what incentive would they have to induce bank runs ?

            If you are interested then see Selgin’s book for some background on Free Banking Theory.


  13. Shane says:

    Any thoughts on targeting George Reismann’s GDR? I think that’d be closer to M*V than GDP.