Book Review: A Random Physicist Takes On Economics

Dr. Jason Smith (publisher of the Information Transfer Economics website) has late released A Random Physicist Takes On Economics. It is a relatively brief, non-technical give-and-take of empirical approaches to economics, as well as is highly critical of mainstream economical approaches. For many of my readers (many of whom also read Jason Smith's work), the criticisms of economic science volition hold upward an interesting read. They are for sure novel, equally they are coming from an outsider's perspective. However, I remove keep closed to reservations almost his approach (which is only briefly summarised inward this book).

Book Description

The mass is relatively brief, alongside the paperback edition having 131 pages. (I remove keep the Kindle ebook version, as well as and hence I cannot refer to page numbers.) The text is non equally good technical, beingness largely without equations (that I retrieve seeing). The text has an introductory department explaining closed to of the hairier theoretical topics he discusses.

Smith's Critiques

In a department labelled "The Critique" (the sections are non numbered), Jason Smith summarises his critiques of economic science equally follows. (The statements below are a straight transcription from the text; I only changed his utilization of "I" to "Smith/He" inward gild to avoid confusion alongside this reviewer.)

• Smith questions the utilization of overly-formal math inward economics.
• He questions the role of human way inward economic science (refers to the random agent concept, discussed below).
• He questions the role of expectations inward economics.
• He questions the agreement of so-called viscid prices inward economics.
• He questions how agents are combined "into an economy." (More specifically, how Dynamic Stochastic General Equilibrium -- DSGE -- theory combines agents into an economical model.)
• He questions the interpretation of the cost machinery (as a aggregation of information).
• He questions the house economic science has relative to the social sciences.

These criticisms overlap those raised yesteryear heterodox economists over recent decades, as well as and hence this aspect of the mass volition remove keep a receptive audience. His analysis approach may hold upward to a greater extent than familiar to those inward the physical as well as applied sciences, as well as and hence the criticisms may brand to a greater extent than sense than heterodox economists for such readers.

Random Agents

The strongest share of the mass is the description of how people acting at random tin sack replicate whatever insights are generated yesteryear utility functions when nosotros human face at consumption behaviour. The thought is that individuals only remove keep a sure amount of coin to brand purchases (a budget constraint), as well as if their choices are taken at random from the laid of possible purchases, an growth inward 1 items cost reduces the number of times it is purchased. This replicates the logic of a demand bend generated yesteryear a utility function.

For readers that are interested inward microeconomics, this should hold upward an interesting discussion. Unfortunately, I am non interested inward microeconomics, equally I encounter the bound from micro demeanour to macro demeanour equally hard to make: existent Earth economical sectors contain a mix of real dissimilar agents, each alongside a differing behaviour. Even if nosotros remove keep a credible description of each story of agent nosotros tin sack obtain relatively arbitrary macro demeanour yesteryear changing the mix of agents.

Pricing

His give-and-take of pricing is curious. He notes a recent study that mainstream models of "sticky prices" are invalid. However, he essentially ignores the decades of empirical work of Keynesian/post-Keynesian economists on how prices are set. For someone who emphasises the importance of empirical grounding of theory, that is a curious omission.

I remove keep express noesis of that post-Keynesian theory, but it emphasises the distinction betwixt flex prices (such equally those laid inward fiscal markets) as well as administered prices. For macro analysis, nosotros are oft most interested inward administered prices, although fossil oil prices are the cardinal "flex price" exception (and mayhap bond yields). For example, it is real hard to believe that the highly influential wage settlements reached yesteryear IG Mettall are the resultant of spousal human relationship leaders as well as employers rolling die behind closed doors.

Economics is non Physics

Although it is fun to hold upward condescending to mainstream economists, closed to of Smith's complaints how practices inward economic science diverge from physics are questionable. 

He wastes the reader's fourth dimension discussing how he was surprised that economic science models remove keep the machinery that expected time to come outcomes influence introduce activity. In physics, the time to come does non unremarkably influence the introduce (I only move on upward alongside trends inward theoretical physics yesteryear watching "The Big Bang Theory," as well as hence that disputation may hold upward incorrect...). That's 1 of the reasons why most people do non scour physics textbooks looking for models to apply to finance as well as economics.

Financial derivatives enquiry is dominated yesteryear both pure as well as applied mathematicians, physicists, as well as engineers (from the theoretical wings of engineering). All of the competent ones are fully aware that existent Earth outcomes tin sack differ from the probability distributions embedded inward their models, as well as all the same all the enquiry I am aware of uses the supposition that the probability distribution of realised outcomes matches the probability distribution used inward determining fair value of fiscal assets. (This is essentially "rational expectations.") One could reasonably debate that nosotros do non know the "true" probability functions describing time to come property prices, as well as and hence nosotros should only cost derivatives randomly. However, doing as well as hence would Pb to a real brusk as well as unhappy career equally a marketplace seat maker. The "rational expectations" hypothesis inward finance is needed to create models that offering us a hazard to operate inward the Earth of fiscal derivatives -- inward other words, they are empirically grounded. Since the supposition industrial plant inward finance, at that spot are no a priori reasons to pass upward its utilization inward economics.

He also raises the number of scope conditions: under what atmospheric condition is a model valid? They are there, only non presented equally they are inward physics -- for reasons I beak over inward a afterward section. For example, economists quite oft piece of work alongside closed economic scheme models: interactions alongside unusual countries are assumed to hold upward insignificant. This manifestly limits the reach of the model.

There are closed to tyke issues. He decries the lack of fault bars inward economic science texts. I dug upward closed to of my old undergraduate electrical technology scientific discipline texts (including my quantum physics text), as well as did non spot a unmarried fault bar inward whatsoever of them. Are nosotros to believe that the measure electrical technology scientific discipline texts are non empirically grounded? In economics, prediction fault analysis is done inward econometrics texts, as well as statistical tests are reported inward macro enquiry (other than a handful of extremely theoretical papers).

Empirical Weakness of DSGE Economics

I am non a fan of DSGE economics, but I remove keep a electrical load that appears to hold upward the exact reverse of Smith's: inward my view, at that spot is no practical way to falsify DSGE macro. This lack of falsifiability is why it remains extremely robust to criticism: both from the exterior as well as fifty-fifty internal reformers.

Meanwhile, Jason Smith takes the stance that DSGE macro has equally good weak predictive powers. He cites studies that exhibit DSGE model predictions performing worse than uncomplicated econometric techniques, or of course, his data transmission economic science techniques. I beak over data transfer afterward inward this review, inward this department I only beak over the underperformance versus uncomplicated models.

Since I stance these empirical comparing studies to hold upward misleading (for reasons discussed here), I volition non delve into them. I volition instead offering a simplified event that roughly captures what those studies found.

In the post-1990 environment, 1 of the best "standard" macro models for predicting inflation inward many developed countries has been to extrapolate previous trends as well as assume that (core) consumer cost inflation volition stay close 2% (0% inward Japan). (Overall CPI inflation oscillates to a greater extent than due to fossil oil prices, but those deviations remove keep been largely transient.) There are closed to notable exceptions, such equally Greece, where euro surface area policy makers managed the feat of recreating depression atmospheric condition inward a welfare state.

In other words, DSGE models remove keep been underperforming a share that only returns 0.02. Although this is funny, almost no economist or fifty-fifty fiscal marketplace seat player cares almost this. Why? 

We are dorsum to Smith's reach conditions. The reach status for the "inflation volition hold upward 2%" model is the electrical flow environs -- characterised yesteryear inflation sticking close 2%. You do non postulate a doctorate inward theoretical physics to encounter that this is a fairly featherbrained situation. Furthermore, such extrapolations of yesteryear demeanour would remove keep failed inward the 1970s, or inward the early on 1990s (the authorities changes). 

The signal of economical models is to offering predictions almost what volition laissez passer on off to changes inward the policy regime. For example, what would hold upward the consequence on inflation yesteryear a large universal guaranteed income scheme? Until the scheme is rolled out nationwide, nosotros remove keep to utilization theory to justice what the macro effects would be. Since uncomplicated models (such equally assuming that inflation volition stick at 2%) cannot offering whatsoever data almost such policy shifts, they are non viewed equally legitimate contenders, as well as and hence their forecasting success over brusk fourth dimension intervals is ignored.

Of course, fiscal marketplace seat participants do desire to brand short-term forecasts of economical variables. I used to follow the inflation-linked market, as well as at that spot is a pocket-size manufacture of analysts who pass their days making short-term inflation forecasts. Needless to say, they did non utilization DSGE models. Instead, they used partial models of the major components of CPI, as well as stuck them together to teach an overall forecast. Although these were accurate models, they were solely at the mercy of policy shifts (a reality of which most were aware).

Information Transfer Economics

The mass only offers a express introduction to what Jason Smith calls information transfer economics. He explains the determination inward "Why No Empirical Work?" To summarise, he prefers to hold off for validation of his ideas inward peer-reviewed journals.

If data transfer economic science is fifty-fifty one-half equally successful equally Smith claims, it would remove keep an incredible commercial value. If I were at my old job, I would remove keep instantly had an analyst taking apart his piece of work equally shortly equally I saw closed to of his model outputs. He has dozens of models that appear to offering highly accurate forecasts for of import economical as well as fiscal variables. Looking for validation inward peer-reviewed journals is curious: if the capitalist organization is an efficient organization for processing information, the commercial success of the techniques should remove keep appeared inside months of their appearance inward the world domain.

I would Federal Reserve annotation that I taught communications systems theory (on which closed to of the data transfer theory is based), a background inward theoretical applied mathematics, equally good equally almost xv years of sense equally a fixed income analyst. In other words, I should hold upward able to attempt the validity of his empirical claims. However, when I attempted to do so, I hitting a rather large route block. The entire data equilibrium theory is only dorsum story for the algorithm he uses to generate forecasts; as well as similar the dorsum story inward old schoolhouse video games similar Pac-Man, it is expendable. However, when I read 1 of his initial papers, the actual algorithm description was only a reference to beginning code inward a reckoner linguistic communication I never worked with, nor had access to. From my perspective, the beginning code was effectively undocumented. I was forced to guess how his algorithm worked. On the footing of that guess, I saw trivial postulate to pursue analysing the algorithm.

I desire to underline that this was purely a justice on my part; I could remove keep made a grievous fault for whatsoever number of reasons. (For one, long-time readers volition Federal Reserve annotation my dour skepticism almost mathematical forecasting models.) Although my website for sure has express traffic, I remove keep a largely self-selected numerate audience. Some of them are probable to hold upward inward a amend seat than myself to justice Jason Smith's claims almost his algorithm.

Concluding Remarks

This mass is interesting, adding all the same to a greater extent than complaints against mainstream economics. However, nosotros postulate to sympathize why DSGE macro is pop earlier edifice expectations that it mightiness disappear. Information transfer economic science offers a challenge to accepted thinking almost economics, but the algorithm functioning needs to hold upward translated into a description that tin sack hold upward analysed yesteryear those accustomed to traditional econometrics.

(c) Brian Romanchuk 2017