Part two of the macroeconomics of forced job loss, promised last week, has been postponed. The reason is a neighborhood 4th-of-July party, at which the buzz was about ill-considered tariffs and whether they would lead to another Great Recession. (A repeat of 2008-09 is always the big fear.) One neighbor, a supplier to auto companies, caught everybody’s attention by reporting that his second-quarter orders, after a robust beginning of the year, had dropped sharply and that he has begun laying-off employees.
Of course, nobody should be surprised that prospects are changing in tariff-affected industries. But the party-goers’ worries were more general, rooted in Donald Trump’s limited grasp of the nature of globally integrated, just-in-time production and apparent innocence of the immense cost of reversing established supply chains. The combination of ignorance on how modern economies work and an unshakable faith in personal intuition is stirring fears that this President could stumble into another hugely costly Great Recession.
Analytic framework. Unlike mainstream macro theory which is badly hampered by its market centricity, the GEM Project provides an analytic framework that captures the essential macrodynamics of extreme instability, most recently on display a decade ago. This space-constrained post will focus on four features of that framework.
First, and most familiar to readers of the GEM Blog, is microfounded meaningful wage rigidity that uniquely motivates causality from adverse nominal demand disturbances to involuntary job loss and proportionally-sized contractions in employment, output, and income. As a result, the GEM Project puts the management of aggregate demand firmly at the center of any effective response to macro instability. The degree to which Trump’s trade war induces economic contraction (from very small dips to calamitous depression) depends on its impact on total spending.
Second, the Project separates adverse nominal demand fluctuations into two types. Stationary demand disturbances (SDD) reflect the contained, temporary weakening of total spending associated with garden-variety recessions. Nonstationary demand disturbances (NND) are the unchecked spending collapses associated with acute instability that threaten depression. (Chapters 5) NDD overwhelms the automatic stabilizers and central-bank purchase of short-term treasury debt that effectively ameliorate SDD.
Third, generalized-exchange modeling has identified an especially important difference between the two classes of spending disturbances that is rooted in prevailing investor/lender perceptions of the macro future. In NDD circumstances, many investors/ lenders become uncertain about the credibility of stabilization authorities’ trend employment/output objective.
The fourth point elaborates on, and provides proper attribution for, NDD uncertainty. In 2009, Nancy Stokey insightfully demonstrated that, as investors/lenders become less certain about macro prospects, simple inaction becomes increasingly rational. Most critically, buyers of financial assets respond to questions about the credibility of the central bank’s real-side objective by moving to the sidelines, waiting for the emergence of a credible floor under prices. A maxim of veteran traders is not to try catching a falling knife. (See Chapter 6 for elaboration on how sufficient inaction fuels collapsing total demand.) In the second contribution, Roger Farmer (2010a, 2010b) has revived an old, again intuitive, idea. Independent of economic fundamentals, investor confidence can influence, and is influenced by, the behavior of prices on asset exchanges. In the simple feedback model used below to capture both Stokey’s and Farmer’s ideas, asset prices are increasingly motivated by self-referential confidence as investor assessment of future macro (mainly aggregate demand) prospects is increasingly governed by uncertainty.
Increasingly uncertain expectations are, by definition, decreasingly governed by priceable risk. In modeling extreme instability, investor anticipations are more complex than just effectively using all available information. A critical macroeconomic issue concerns the degree to which effective probabilities can be assigned to future states of the macroeconomy. Frank Knight (1921) famously analyzed the distinction between priceable risk, requiring informed priors on outcome likelihoods, and unpriceable uncertainty, arising from unknown probability distributions.
Simple model. In modeling NDD, the GEM Project focuses on investment outlays, the most volatile component of total spending. The basic model is constructed in Chapter 6:
I(t)=ƒ[Ƈ(t)И(t), (1-Ƈ(t))₡(t)], such that ΔI/ΔИ>0, ΔI/Δ₡>0, 0<Ƈ<1,
where I denotes nominal investment outlays on capital goods, construction, and software at time t; И represents discounted, inflation-adjusted expectations of future profits rooted in economic fundamentals; ₡ denotes investor confidence, the Project’s version of Keynes’s animal spirits; and Ƈ calibrates investor perceptions of the trend real-side credibility of stabilization authorities.
Ƈ crucially mediates the relative influence of economic fundamentals and self-referential confidence in investment spending. When credibility is high, investors’ macro expectations are governed by generally agreed-upon probabilities; and investment spending is almost wholly driven by profit fundamentals. As credibility erodes, however, the investors become more uncertain, causing rational inaction and investment spending to be increasingly driven by faltering confidence, producing gathering weakness in total demand. Containing NDD requires relatively aggressive use of the central-bank balance sheet, supporting the crisis-imposed roles of buyer and guarantor of last resort in downward-spiraling markets, in order to restore effective recycling of saving into spending sufficiently to reverse the contraction of nominal spending.
In late 2008 and early 2009, Ƈ played a central role in the brewing spending collapse. Investors and lenders clearly became uncertain about the reliability of trend real-side stabilization goals, becoming rationally inactive and pushing asset prices and total demand into downward spirals. This post has insufficient room to rationally model Ƈ. (See Chapters 6 and 10.) Perhaps it will suffice to consider the model’s counter-factual case. If investors/lenders in 2008-09 had been confident with respect to the Fed’s capacity and will to sustain trend aggregate demand in the face of the disruption associated with expected subprime and exotic residential-mortgage default, private-sector decision-making would have continued to be guided by profit-seeking rooted in economic fundamentals. (Overall fundamentals of the U.S. economy were little altered by the mortgage crisis.) The equity-price decline would have been much less than the 50% that actually occurred. There would have been no NDD contraction in total spending, output, and employment.
What to do? Within the context provided by the GEM Project, stabilization policymakers always have two basic, not mutually exclusive, strategies to guide their efforts to prevent a repeat, or worse, of the massive welfare loss of the Great Recession:
- Prevent future shocks that can damage the economy’s capacity to recycle saving into investment spending.
- If such disruptions do occur, prevent their propagation by effectively intervening to halt and reverse contracting nominal demand.
If someone knows how to convince Donald Trump that he is naïve about the nature and consequences of ill-considered tariffs, he or she should step forward and prevent a risky trade war. In the absence of such an effective crusader against ignorance, stabilization policymakers must think hard both about how to identify when their real-side credibility is faltering and how best to execute the second strategy.
Blog Type: Policy/Topical Saint Joseph, Michigan