No, The Fed Is Keeping Rates Low

As is forthwith good known, Ben Bernanke has only launched a weblog (welcome!) at the Brookings Institution. His get-go analysis ship was "Yes, Interest Rates Are Artificial." Correspondingly, I remember his request hither is brilliant. My concern almost the quip regarding the coin render partly raises a somewhat theoretical argument*, but I believe that his declaration does non reverberate how monetary operations are implemented. I may necktie this inward amongst my series of posts on authorities operations in a subsequently article..

Long-Term Equilibrium Rates

The key request of his declaration is inside this paragraph.

If y'all asked the someone inward the street, “Why are involvement rates in addition to then low?”, he or she would probable respond that the Fed is keeping them low. That’s truthful entirely inward a really narrow sense. The Fed does, of course, laid the benchmark nominal short-term involvement rate. The Fed’s policies are also the primary determinant of inflation in addition to inflation expectations over the longer term, in addition to inflation trends impact involvement rates, every bit the figure inward a higher house shows. But what matters most for the economic scheme is the real, or inflation-adjusted, involvement charge per unit of measurement (the market, or nominal, involvement charge per unit of measurement minus the inflation rate). The existent involvement charge per unit of measurement is most relevant for upper-case missive of the alphabet investment decisions, for example. The Fed’s powerfulness to impact existent rates of return, specially longer-term existent rates, is transitory in addition to limited. Except inward the curt run, existent involvement rates are determined past times a broad arrive at of economical factors, including prospects for economical growth—not past times the Fed.

My electrical load is that the "the someone on the street" (or fifty-fifty "the someone on Wall Street") is truthful inward a broad sense, non a narrow sense.

The reaction from some post-Keynesian bloggers was fairly critical, such every bit Ramanan's answer inward "Michal Kalecki on the Great Moderation" (and Kalecki himself was writing inward 1943).

However, I desire to follow the editorial delineate taken past times "circuit" inward "Ben Bernanke in addition to the natural charge per unit of measurement of interest."

Now, I realize that the equilibrium existent charge per unit of measurement is unobservable in addition to varies through time, which way it's discipline to uncertainty. However, nosotros could say the same thing almost the concept of potential output, yet few would deny it is a useful concept.
In fact, most people are aware of the concept of "output gap", the difference betwixt potential output in addition to actual output. The corollary concept for the existent involvement charge per unit of measurement is the "interest charge per unit of measurement gap", the divergence of the actual policy charge per unit of measurement from the existent equilibrium rate.

I am willing to grant that the "equilibrium existent rate" is an unobservable model variable. My declaration is that Ben Bernanke even in addition to then cannot depict the conclusions he does fifty-fifty if this is true.

Theoretical Frameworks Versus Mathematical Models

MD Bernanke is basically writing inward the same way he would address the students that he used to teach, in addition to next touchstone mainstream logic. I would paraphrase his declaration inward this way (although I am extending his logic inward what I believe is a reasonable fashion).

• We assume an economical framework inward which the existent charge per unit of measurement of involvement has a strong outcome on behaviour.
• This framework is converted to a simplified mathematical (typically using log-linearisation), in addition to that simplified model is jibe to existent earth data.
• The jibe is statistically acceptable (although it would non hold upwards able to predict things similar the fiscal crisis, every bit that crisis was allegedly the resultant of a non-forecastable "productivity shock"). Therefore, nosotros tin debate that the model is "correct", statistically speaking.
• Since the charge per unit of measurement of inflation was "acceptable" to policymakers during the post-1982 era, the policy involvement charge per unit of measurement was nigh the model-predicted "equilibrium" grade (which changes over time). This holds because if had been held elsewhere, the model predicts that at that spot would receive got been a large divergence of the inflation charge per unit of measurement (higher or lower).
• Therefore, policymakers were next this equilibrium, in addition to and then it was non a witting alternative to select almost depression existent involvement rates.

Missing Ingredient - Model Error

The mainstream economical modelling framework implicitly assumes at that spot are ii sources of mistake (or risk) inside macroeconomic modelling.

1. The parameters of the fitted model are estimated incorrectly. (These are the "structural parameters" of DSGE models.)
2. The set down of the organisation (such as, what is the output gap inward March 2015?) is estimated incorrectly.

They purpose was appears to hold upwards highly sophisticated statistical analysis (typically Bayes' Theorem based approaches) to demo that these estimation errors are acceptable. This is precisely the type of logic used past times Optimal Control theorists inward the 1960s. As I discussed previously, optimal command was abandoned past times engineers after it killed a evidence pilot. From the request of reckon of a command systems theorist (which I am), mainstream economists are trapped inward a 1960s fourth dimension warp.

What is missed is the tertiary type of adventure - model risk. 

We practise non know the exact model of whatsoever applied scientific discipline system. For example, inward aerospace, the airframe bends, in addition to nosotros cannot promise to capture the higher harmonics of the oscillations. But if nosotros desire sophisticated applied scientific discipline systems, nosotros cannot arrive at upwards on modelling in addition to collapse into a funk. Instead, nosotros demand to adopt approaches that explicitly contain model errors, which is an approach named "robust control".**

Once nosotros select model adventure into account, things larn interesting.

Assume that nosotros receive got a prissy mainstream model, which fits the information acceptably. As I argued earlier, the realised policy charge per unit of measurement would receive got been unopen to the equilibrium rate.

Then imagine that at that spot is a perturbed version of the model, which has the same dynamics every bit this master copy model, nether the supposition that the "fixed parameters" are the same every bit the original, in addition to the policy charge per unit of measurement is really unopen to the "equilibrium existent rate" inward the get-go model. (We tin infer that the "equilibrium rate" exists inside the minute model, in addition to is the same every bit the first.)

However, the difference is that for this perturbed model, the outcome of the existent policy charge per unit of measurement is less than the master copy model if nosotros movement away from equilibrium. This mightiness hold upwards the resultant of inertia inward entities' economical deportment (for example, the utility business office penalises changes changes inward production levels versus their levels inward the previous accounting period). In other words, this model has a lower sensitivity to the existent charge per unit of measurement than the baseline model.

This novel model could hold upwards jibe the information inward the same way, in addition to generate a similar character of fit. (It may hold upwards that the "best jibe parameters" are slightly different, to select into job organisation human relationship the pocket-size deviations of the realised policy charge per unit of measurement from the model-predicted equilibrium.)

Taking this novel model into account, nosotros tin no longer depict the conclusion that Ben Bernanke does. If the economic scheme is non every bit sensitive to the policy charge per unit of measurement every bit he assumes, it could easily hold upwards that the equilibrium nominal policy charge per unit of measurement inward the USA right forthwith is 2%, non the 0% that he is taken every bit given. The Fed may receive got been deliberately keeping the existent charge per unit of measurement of involvement several hundred set down points besides low, but was non plenty to receive got an appreciable outcome on inflation. In other words, this was a policy alternative past times the Fed.

Since these ii models are essentially indistinguishable unless the Fed deliberately engineers a huge divergence from the model-predicted existent rate, no amount of statistical obfuscation tin tell us which is correct. It turns into a sentence call. In my opinion, the fact that the estimated "equilibrium existent rate" swings approximately in addition to then much is a sign that existent rates are non that powerful a driver for the economy. But obviously, opinions differ.

Concluding Remarks

It is prissy to encounter Ben Bernanke laid upwards a blog, every bit he should hold upwards able to clearly enunciate the mainstream view. But nosotros demand to delve beyond the high-level academic generalities to encounter what the of import distinctions betwixt the mainstream in addition to heterodox economical reckon are.

Footnotes:

* I am non going to dredge upwards the discussion "exogenous" here.

** There has been some attempts to innovate robust command into economics, such every bit inward the mass "Robustness" past times Hansen in addition to Sargent. However, they adopted the worst possible way of looking at robust control, which was a formulation relying upon game theory. Within this approach, model mistake disappears, it's only that y'all are contending amongst "malicious disturbances". This is mathematically equivalent to what I reckon every bit the proper way of looking at robust command entirely inward the instance of linear systems. My feeling is that the economists honey game theory in addition to adopted that methodology every bit a result, in addition to completely lost sight of the number of model risk.

(c) Brian Romanchuk 2015