tag:blogger.com,1999:blog-5908830827135060852.post7743377138977771513..comments2021-03-08T09:33:07.680-05:00Comments on Bond Economics: The Curious Household Accounting Of DSGE ModelsBrian Romanchukhttp://www.blogger.com/profile/02699198289421951151noreply@blogger.comBlogger14125tag:blogger.com,1999:blog-5908830827135060852.post-12332875026422425962018-03-27T11:27:19.852-04:002018-03-27T11:27:19.852-04:00https://goo.gl/images/9fRFcShttps://goo.gl/images/9fRFcSBrian Romanchukhttps://www.blogger.com/profile/02699198289421951151noreply@blogger.comtag:blogger.com,1999:blog-5908830827135060852.post-9107017619995623082018-03-27T10:50:19.308-04:002018-03-27T10:50:19.308-04:00The curious non-existence of profit in economics
C...The curious non-existence of profit in economics<br />Comment on Brian Romanchuk on “The Curious Household Accounting Of DSGE Models”<br /><br />Krugman once put it in a nutshell: “most of what I and many others do is sorta-kinda neoclassical because it takes the maximization-and-equilibrium world as a starting point.”<br /><br />Krugman, of course, is an idiot. But he is not the only one, just the opposite, he is the mouthpiece of the stupid majority. To this basket of deplorables belong also Lars Ljungqvist, Thomas Sargent, the rest of DSGEers, Brian Romanchuk, Roger Sparks, Nick Rowe, and the rest of tireless nonsense bloggers.<br /><br />What unites these folks is scientific incompetence, more specifically, the inability to realize that maximization-and-equilibrium has always been and will always be a methodologically inadmissible starting point. To recall, these are the verbalized neo-Walrasian axioms:<br /><br />HC1. There exist economic agents.<br />HC2. Agents have preferences over outcomes.<br />HC3. Agents independently optimize subject to constraints.<br />HC4. Choices are made in interrelated markets.<br />HC5. Agents have full relevant knowledge.<br />HC6. Observable economic outcomes are coordinated, so they must be discussed with reference to equilibrium states. (Weintraub)<br /><br />These premises contain three plain NONENTITIES (constrained optimization, rational expectations, equilibrium) and therefore are forever unacceptable. Being incompetent scientists, though, most economists swallowed this inane stuff hook, line and sinker from Jevons/Walras/Menger onward to DSGE.<br /><br />Every theory/model that contains just one NONENTITY is scientifically worthless.<br /><br />What has to be realized in addition is that false premises require a pigtail of false auxiliary assumptions. So, in order to be applicable, constrained optimization requires the auxiliary assumption of a production functions with decreasing returns. Thus, a silly behavioral assumption determines the physical properties of production in the upside-down world of standard economics.<br /><br />One auxiliary assumption that is clearly at odds with reality is zero profits.<br /><br />“The zero profit attributed to perfect competition does not arise from perfect competition at all, but merely from the assumption that aggregate profit is zero.” (Murad)<br /><br />“… since it is impossible to have an economy where everyone is making profits. Aggregate profit for an entire (closed) economy must be zero, hence if any firm is making profits, some other firm must be making losses.” (Boland)<br /><br />“Wherever entrepreneurs make profits … they expand production; wherever they incur losses, production is contracted. In equilibrium therefore, there are neither profits nor losses. Walras thus created the abstraction of the zero-profit entrepreneur under perfect competition.” (Niehans)<br /><br />“Profit theory has long been regarded as one of the more unsatisfactory branches of economics. . . . One reason for this is that economists have not asked the right questions about profit.” (Murad)<br /><br />In fact, economists do not understand since 200+ years what profit is.#1 This includes Walrasians, Keynesians, Marxians, Austrians, Pluralists, DSGEers, MMTers and, last but not least, Brian Romanchuk.<br /><br />Lars Ljungqvist’s and Thomas Sargent’s DSGE is proto-scientific dreck. Nobody with more than two brain cells needs three lengthy posts to arrive at this conclusion.<br /><br />Egmont Kakarot-Handtke<br /><br />#1 For details see cross-references Profit<br />http://axecorg.blogspot.de/2015/03/profit-cross-references.htmlAXEC / E.K-Hhttps://www.blogger.com/profile/10402274109039114416noreply@blogger.comtag:blogger.com,1999:blog-5908830827135060852.post-62547441514763757262018-03-26T14:56:40.879-04:002018-03-26T14:56:40.879-04:00"To say that published DSGE treatments of mul..."To say that published DSGE treatments of multi-sector models are muddled is an understatement."<br /><br />I am but a mere lowly student of this area but it seems to me that DSGE modelling is a complete contrivance with questionable logic.<br /><br />Firstly, the model is specified and defined. It is said to be based on GE but this hardly appears to be the case - it would seem it is more like a partial equilibrium approach. As Brian says, the major mathematical operation is the use of optimization algorithms. The model is linearized to remove cyclical elements. Parameters are estimated (guesstimated) and then the model equations are calibrated and massaged to fit the data.<br /><br />All this seems a patchwork of steroidal techniques and hardly elegant.<br /><br />Henry RechAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-5908830827135060852.post-82043298356900224512018-03-26T14:39:32.740-04:002018-03-26T14:39:32.740-04:00The phrase "we owe it to ourselves" is t...The phrase "we owe it to ourselves" is true only if the household sector is NOT divided into three segments: wealthy, working class, and poor. In that case the wealthy own equity and debt instruments which represent future cash flow obligations of the working class. The government can be seen as another type of firm which intermediates between the wealthy and working class households. So government policy has much to do with the evolving structure of the financial float in a context where bailouts may occur, debts may be discharged in bankruptcy, financial assets may be zombie items or written off, and government may run a surplus or deficit.Joe Leotehttps://www.blogger.com/profile/01292763300917387201noreply@blogger.comtag:blogger.com,1999:blog-5908830827135060852.post-76375567512077111372018-03-26T13:17:10.872-04:002018-03-26T13:17:10.872-04:00Thanks Brian. I think I have it now.
The product...Thanks Brian. I think I have it now. <br /><br />The production equation assumes no beginning capital. As a result of that assumption, at end of any time period, \(k(t+1) = k(t).\) <br /><br />If there was a beginning capital quantity, we would have at end of any time period (t) \(k(t + 1) = k(t + 0) + k(t).\)<br /><br />I think that is the correct interpretation.Roger Sparkshttps://www.blogger.com/profile/01734503500078064208noreply@blogger.comtag:blogger.com,1999:blog-5908830827135060852.post-55496174441033652062018-03-26T12:37:03.048-04:002018-03-26T12:37:03.048-04:00Time notation: all series are defined for time=0, ...Time notation: all series are defined for time=0, 1, 2, 3, ...<br />x(t) refers to the variable at the current time; for example x(0) refers to the value of x at time zero.<br />x(t+1) refers to the value at the next time point, So if t=0, x(t+1) is the value at time 1. This is used to tie capital levels and debt levels to the previous period.Brian Romanchukhttps://www.blogger.com/profile/02699198289421951151noreply@blogger.comtag:blogger.com,1999:blog-5908830827135060852.post-376468398812976332018-03-26T11:26:32.125-04:002018-03-26T11:26:32.125-04:00Thanks Nick. The perspective you present is consid...Thanks Nick. The perspective you present is considerably different from my beginning perspective.<br /><br />One thing: I think the time notation \((t)\) represents a data accumulation that occurs over a time period. In contrast, the time notation \((t+1)\) would be the single point reading at time point one.<br /><br />If that would be the correct notation assignment, I might still quible with the equation over the apparent useage of time intervals.Roger Sparkshttps://www.blogger.com/profile/01734503500078064208noreply@blogger.comtag:blogger.com,1999:blog-5908830827135060852.post-11861553385575540812018-03-26T09:57:12.304-04:002018-03-26T09:57:12.304-04:00The lack of money appearing is an issue (and raise...The lack of money appearing is an issue (and raised the ire of a certain commenter...), and the zero "accounting profits" made my points less interesting. I was going to get to the governmental side, and discuss some other models as well, in which case the results are more interesting.<br /><br />I like L&S since they start off close to the optimal control (and my Ph.D. is in control systems). Other treatments buried the optimal control roots, and the mathematics just mystified me. (Gali's text in particular.)<br /><br />Unfortunately, my consulting duties call, so I would have to get back to any other points later...Brian Romanchukhttps://www.blogger.com/profile/02699198289421951151noreply@blogger.comtag:blogger.com,1999:blog-5908830827135060852.post-64442003303094434812018-03-26T08:58:00.639-04:002018-03-26T08:58:00.639-04:00Brian: I haven't read L&S. It's not re...Brian: I haven't read L&S. It's not really my cup of tea. I'm impressed you are having a crack at it. I think you are mostly getting the intuition.<br /><br />There's always hidden (or non-obvious) assumptions in math models. The trick is to figure out which ones might matter.<br /><br />I don't see anything that could prevent government debt being negative in this model. Instead of the government owing the households, the households owe the government. And with lump sum taxes and infinitely-lived representative household, it wouldn't matter anyway, whether the right pocket owes the left pocket or vice versa. Lump sum taxes could be negative too (i.e. transfer payments).<br /><br />But there's no money in this model (yet). We have to think of bonds being promises to pay goods. Wages and capital rentals and taxes are also paid in goods.<br /><br />If firms own the capital, and pay the rents to themselves, then those rents would be "profits" in the accounting sense of the word, and those "profits" would appear in the household budget instead of rents. But since the two are equal (constant returns to scale), it wouldn't make any difference to the model. Households own the firms, just like they "own" the government (they owe its debt to themselves). The whole institutional structure is irrelevant in this model, except for: government spending (which is purely wasteful, since it does not appear in the household's utility function); and the distorting taxes.Nick Rowehttps://www.blogger.com/profile/04982579343160429422noreply@blogger.comtag:blogger.com,1999:blog-5908830827135060852.post-76707629351968688722018-03-26T06:47:41.845-04:002018-03-26T06:47:41.845-04:00OK, will have to think about your comments. They j...OK, will have to think about your comments. They jibe with the bulk of my analysis. The issue I see is that the taxes only matter if we insist that government debt cannot be negative; the households can say, “sure, we’ll pay your taxes - just keep lending us the money to do so.” Mathematically, that is what the discussion of the optimal solution should revolve around. <br /><br />The case with positive profits would perhaps be the more interesting case; it is what I was used to. This article was a first draft, and I discovered my conclusions were weaker for this model than the other case (the one in the text by Gali) only after I finished my write up.<br /><br />There’s presumably hidden mathematical assumptions (perhaps discussed in the previous 614 pages) that I am missing, that precludes my solution technique. My academic-style complaint is that applied mathematics can’t bury critical assumptions like that.Brian Romanchukhttps://www.blogger.com/profile/02699198289421951151noreply@blogger.comtag:blogger.com,1999:blog-5908830827135060852.post-51331130776780755042018-03-26T00:31:09.236-04:002018-03-26T00:31:09.236-04:00Roger: no.
The first equation in your comment is ...Roger: no.<br /><br />The first equation in your comment is an engineering relationship. It tells us that the level of output depends on the level of employment of labour and kapital (machines), and output can be either consumed by households, or government, or used to increase the stock of machines.<br /><br />The production function F(t,k(t),n(t)) is not specified. It does not need to be specified. But y=k^0.3xn^0.7 would be a numerical example that fits the model.<br /><br />The second equation adds that all output (income) gets paid out as wages and machine rentals, and gets used as per above.<br /><br />r is rental paid per machine per period, just like the wage paid for labour. delta is the fraction of machines that explode per period. There is no inflation in this model, because there is no money (yet). Think of it as a barter economy.Nick Rowehttps://www.blogger.com/profile/04982579343160429422noreply@blogger.comtag:blogger.com,1999:blog-5908830827135060852.post-63293651603579557812018-03-25T23:25:13.073-04:002018-03-25T23:25:13.073-04:00I remain frustrated with the math presented here. ...I remain frustrated with the math presented here. I am focusing on equation 16.2.3. \[c(t) + g(t) + k(t+1) = F(t, k(t), n(t)) + (1-\delta)k(t)\]<br /><br />I am not well versed in function notation but, based on what I can learn from Wikipedia and other sources, funtion notation is likened to a black box from which we expect to see emerge a prescribed transformation of variable inputs.<br /><br />If the 'black box' analogy is correct, the actual production equation does not appear until later following substitutions into eq 16.2.3. The result, which would be the 'production equation', is \[c(t) + g(t) + k(t+1) = r(t)k(t) + w(t)n(t) + (1-\delta) k(t).\]<br /><br />Terms r and delta are both related to changes in value of capital, but presented in different contexts. Hence, we have a single variable, inflation, influencing the production equation in two different but related ways. <br /><br />So what is the final usefulness of this equation? The combined effect of double entry seems to mostly offset. The equation remains very confusing to me.<br /><br />Roger Sparkshttps://www.blogger.com/profile/01734503500078064208noreply@blogger.comtag:blogger.com,1999:blog-5908830827135060852.post-86387757657412058872018-03-25T22:56:47.623-04:002018-03-25T22:56:47.623-04:00Actually, you could pretty much delete firms from ...Actually, you could pretty much delete firms from this model, assume the representative household produces output using its own labour and machines, and it wouldn't make any difference. It's just a way to get distorting taxes into the model, because if would be hard for the government to observe wages and machine rentals, to tax them, in a self-employed economy.Nick Rowehttps://www.blogger.com/profile/04982579343160429422noreply@blogger.comtag:blogger.com,1999:blog-5908830827135060852.post-27507036673432211952018-03-25T22:13:25.309-04:002018-03-25T22:13:25.309-04:00Brian: I skimmed this post, and your previous post...Brian: I skimmed this post, and your previous post.<br />As you probably know, I don't do math. But here's a couple of comments that might help with the intuition:<br /><br />Suppose households own all the machines ("capital"). Firms own nothing. Firms hire labour and machines from households, and pay them by the hour. If individual firms maximise profits, taking prices, wages, and the machine rental rate as given, wage and rental rate equal marginal products of labour and machines. And given constant returns to scale, those payments to households for renting labour and machines must add up to total output, so profits must be zero. (And the number of firms is irrelevant, as long as it's large enough to ensure none has market power over prices and wages and rentals.)<br /><br />If you instead assume decreasing returns to scale (like your square root example), then firms would earn positive profits, and you would need to add those profits into the household budget constraint. And the number of firms would matter (with free entry the number of firms would explode to infinity). And it would be a very different model.<br /><br />If taxes were lump sum, the stock of government bonds would be irrelevant in this model. Standard Ricardian Equivalence result for an infinitely-lived household. "We owe it to ourselves". The only reason the stock of bonds matters in this model is that taxes are distorting, so will affect the household's consumption/leisure and intertemporal consumption/saving choices.<br /><br />Dunno if this helps.Nick Rowehttps://www.blogger.com/profile/04982579343160429422noreply@blogger.com