Monthly Archives: October 2022

Stealth Hiring Quotas Via AI

24 Monday Oct 2022

Posted by in Discrimination, Diversity, Quotas, Uncategorized

≈ Leave a comment

Tags

, , , , , , , , , , , , , , , ,

Hiring quotas are of questionable legal status, but for several years, some large companies have been adopting quota-like “targets” under the banner of Diversity, Equity and Inclusion (DEI) initiatives. Many of these so-called targets apply to the placement of minority candidates into “leadership positions”, and some targets may apply more broadly. Explicit quotas have long been viewed negatively by the public. Quotas have also been proscribed under most circumstances by the Supreme Court, and the EEOC’s Compliance Manual still includes rigid limits on when the setting of minority hiring “goals” is permissible.

Yet large employers seem to prefer the legal risks posed by aggressive DEI policies to the risk of lawsuits by minority interests, unrest among minority employees and “woke” activists, and “disparate impact” inquiries by the EEOC. Now, as Stewart Baker writes in a post over at the Volokh Conspiracy, employers have a new way of improving — or even eliminating — the tradeoff they face between these risks: “stealth quotas” delivered via artificial intelligence (AI) decisioning tools.

Skynet Smiles

A few years ago I discussed the extensive use of algorithms to guide a range of decisions in “Behold Our Algorithmic Overlords“. There, I wrote:

“Imagine a world in which all the information you see is selected by algorithm. In addition, your success in the labor market is determined by algorithm. Your college admission and financial aid decisions are determined by algorithm. Credit applications are decisioned by algorithm. The prioritization you are assigned for various health care treatments is determined by algorithm. The list could go on and on, but many of these ‘use-cases’ are already happening to one extent or another.”

That post dealt primarily with the use of algorithms by large tech companies to suppress information and censor certain viewpoints, a danger still of great concern. However, the use of AI to impose de facto quotas in hiring is a phenomenon that will unequivocally reduce the efficiency of the labor market. But exactly how does this mechanism work to the satisfaction of employers?

Machine Learning

As Baker explains, AI algorithms are “trained” to find optimal solutions to problems via machine learning techniques, such as neural networks, applied to large data sets. These techniques are are not as straightforward as more traditional modeling approaches such as linear regression, which more readily lend themselves to intuitive interpretation of model results. Baker uses the example of lung x-rays showing varying degrees of abnormalities, which range from the appearance of obvious masses in the lungs to apparently clear lungs. Machine learning algorithms sometimes accurately predict the development of lung cancer in individuals based on clues that are completely non-obvious to expert evaluators. This, I believe, is a great application of the technology. It’s too bad that the intuition behind many such algorithmic decisions are often impossible to discern. And the application of AI decisioning to social problems is troubling, not least because it necessarily reduces the richness of individual qualities to a set of data points, and in many cases, defines individuals based on group membership.

When it comes to hiring decisions, an AI algorithm can be trained to select the “best” candidate for a position based on all encodable information available to the employer, but the selection might not align with a hiring manager’s expectations, and it might be impossible to explain the reasons for the choice to the manager. Still, giving the AI algorithm the benefit of the doubt, it would tend to make optimal candidate selections across reasonably large sets of similar, open positions.

Algorithmic Bias

A major issue with respect to these algorithms has been called “algorithmic bias”. Here, I limit the discussion to hiring decisions. Ironically, “bias” in this context is a rather slanted description, but what’s meant is that the algorithms tend to select fewer candidates from “protected classes” than their proportionate shares of the general population. This is more along the lines of so-called “disparate impact”, as opposed to “bias” in the statistical sense. Baker discusses the attacks this has provoked against algorithmic decision techniques. In fact, a privacy bill is pending before Congress containing provisions to address “AI bias” called the American Data Privacy and Protection Act (ADPPA). Baker is highly skeptical of claims regarding AI bias both because he believes they have little substance and because “bias” probably means that AIs sometimes make decisions that don’t please DEI activists. Baker elaborates on these developments:

“The ADPPA was embraced almost unanimously by Republicans as well as Democrats on the House energy and commerce committee; it has stalled a bit, but still stands the best chance of enactment of any privacy bill in a decade (its supporters hope to push it through in a lame-duck session). The second is part of the AI Bill of Rights released last week by the Biden White House.”

What the hell are the Republicans thinking? Whether or not it becomes a matter of law, misplaced concern about AI bias can be addressed in a practical sense by introducing the “right” constraints to the algorithm, such as a set of aggregate targets for hiring across pools of minority and non-minority job candidates. Then, the algorithm still optimizes, but the constraints impinge on the selections. The results are still “optimal”, but in a more restricted sense.

Stealth Quotas

As Baker says, these constrains on algorithmic tools would constitute a way of imposing quotas on hiring that employers won’t really have to explain to anyone. That’s because: 1) the decisioning rationale is so obtuse that it can’t readily be explained; and 2) the decisions are perceived as “fair” in the aggregate due to the absence of disparate impacts. As to #1, however, the vendors who create hiring algorithms, and specific details regarding algorithm development, might well be subject to regulatory scrutiny. In the end, the chief concern of these regulators is the absence of disparate impacts, which is cinched by #2.

About a month ago I posted about the EEOC’s outrageous and illegal enforcement of disparate impact liability. Should I welcome AI interventions because they’ll probably limit the number of enforcement actions against employers by the EEOC? After all, there is great benefit in avoiding as much of the rigamarole of regulatory challenges as possible. Nonetheless, as a constraint on hiring, quotas necessarily reduce productivity. By adopting quotas, either explicitly or via AI, the employer foregoes the opportunity to select the best candidate from the full population for a certain share of open positions, and instead limits the pool to narrow demographics.

Demographics are dynamic, and therefore stealth quotas must be dynamic to continue to meet the demands of zero disparate impact. But what happens as an increasing share of the population is of mixed race? Do all mixed race individuals receive protected status indefinitely, gaining preferences via algorithm? Does one’s protected status depend solely upon self-identification of racial, ethnic, or gender identity?

For that matter, do Asians receive hiring preferences? Sometimes they are excluded from so-called protected status because, as a minority, they have been “too successful”. Then, for example, there are issues such as the classification of Hispanics of European origin, who are likely to help fill quotas that are really intended for Hispanics of non-European descent.

Because self-identity has become so critical, quotas present massive opportunities for fraud. Furthermore, quotas often put minority candidates into positions at which they are less likely to be successful, with damaging long-term consequences to both the employer and the minority candidate. And of course there should remain deep concern about the way quotas violate the constitutional guarantee of equal protection to many job applicants.

The acceptance of AI hiring algorithms in the business community is likely to depend on the nature of the positions to be filled, especially when they require highly technical skills and/or the pool of candidates is limited. Of course, there can be tensions between hiring managers and human resources staff over issues like screening job candidates, but HR organizations are typically charged with spearheading DEI initiatives. They will be only too eager to adopt algorithmic selection and stealth quotas for many positions and will probably succeed, whether hiring departments like it or not.

The Death of Merit

Unfortunately, quotas are socially counter-productive, and they are not a good way around the dilemma posed by the EEOC’s aggressive enforcement of disparate impact liability. The latter can only be solved only when Congress acts to more precisely define the bounds of illegal discrimination in hiring. Meanwhile, stealth quotas cede control over important business decisions to external vendors selling algorithms that are often unfathomable. Quotas discard judgements as to relevant skills in favor of awarding jobs based on essentially superficial characteristics. This creates an unnecessary burden on producers, even if it goes unrecognized by those very firms and is self-inflicted. Even worse, once these algorithms and stealth quotas are in place, they are likely to become heavily regulated and manipulated in order to achieve political goals.

Baker sums up a most fundamental objection to quotas thusly:

“Most Americans recognize that there are large demographic disparities in our society, and they are willing to believe that discrimination has played a role in causing the differences. But addressing disparities with group remedies like quotas runs counter to a deep-seated belief that people are, and should be, judged as individuals. Put another way, given a choice between fairness to individuals and fairness on a group basis, Americans choose individual fairness. They condemn racism precisely for its refusal to treat people as individuals, and they resist remedies grounded in race or gender for the same reason.”

Quotas, and stealth quotas, substitute overt discrimination against individuals in non-protected classes, and sometimes against individuals in protected classes as well, for the imagined sin of a disparate impact that might occur when the best candidate is hired for a job. AI algorithms with protection against “algorithmic bias” don’t satisfy this objection. In fact, the lack of accountability inherent in this kind of hiring solution makes it far worse than the status quo.

Hurricane—Warming Link Is All Model, No Data

18 Tuesday Oct 2022

Posted by in Climate science, Hurricanes, Uncategorized

≈ Leave a comment

Tags

, , , , , , , , , , , , , , , , , , , ,

There was deep disappointment among political opponents of Florida Governor Ron DeSantis at their inability to pin blame on him for Hurricane Ian’s destruction. It was a terrible hurricane, but they so wanted it to be “Hurricane Hitler”, as Glenn Reynolds noted with tongue in cheek. That just didn’t work out for them, given DeSantis’ competent performance in marshaling resources for aid and cleanup from the storm. Their last ditch refuge was to condemn DeSantis for dismissing the connection they presume to exist between climate change and hurricane frequency and intensity. That criticism didn’t seem to stick, however, and it shouldn’t.

There is no linkage to climate change in actual data on tropical cyclones. It is a myth. Yes, models of hurricane activity have been constructed that embed assumptions leading to predictions of more hurricanes, and more intense hurricanes, as temperatures rise. But these are models constructed as simplified representations of hurricane development. The following quote from the climate modelers at the Geophysical Fluid Dynamics Laboratory (GFDL) (a division of the National Oceanic and Atmospheric Administration (NOAA)) is straightforward on this point (emphases are mine):

“Through research, GFDL scientists have concluded that it is premature to attribute past changes in hurricane activity to greenhouse warming, although simulated hurricanes tend to be more intense in a warmer climate. Other climate changes related to greenhouse warming, such as increases in vertical wind shear over the Caribbean, lead to fewer yet more intense hurricanes in the GFDL model projections for the late 21st century.

Models typically are said to be “calibrated” to historical data, but no one should take much comfort in that. As a long-time econometric modeler myself, I can say without reservation that such assurances are flimsy, especially with respect to “toy models” containing parameters that aren’t directly observable in the available data. In such a context, a modeler can take advantage of tremendous latitude in choosing parameters to include, sensitivities to assume for unknowns or unmeasured relationships, and historical samples for use in “calibration”. Sad to say, modelers can make these models do just about anything they want. The cautious approach to claims about model implications is a credit to GFDL.

Before I get to the evidence on hurricanes, it’s worth remembering that the entire edifice of climate alarmism relies not just on the temperature record, but on models based on other assumptions about the sensitivity of temperatures to CO2 concentration. The models relied upon to generate catastrophic warming assume very high sensitivity, and those models have a very poor track record of prediction. Estimates of sensitivity are highly uncertain, and this article cites research indicating that the IPCC’s assumptions about sensitivity are about 50% too high. And this article reviews recent findings that carbon sensitivity is even lower, about one-third of what many climate models assume. In addition, this research finds that sensitivities are nearly impossible to estimate from historical data with any precision because the record is plagued by different sources and types of atmospheric forcings, accompanying aerosol effects on climate, and differing half-lives of various greenhouse gases. If sensitivities are as low as discussed at the links above, it means that predictions of warming have been grossly exaggerated.

The evidence that hurricanes have become more frequent or severe, or that they now intensify more rapidly, is basically nonexistent. Ryan Maue and Roger Pielke Jr. of the University of Colorado have both researched hurricanes extensively for many years. They described their compilation of data on land-falling hurricanes in this Forbes piece in 2020. They point out that hurricane activity in older data is much more likely to be missing and undercounted, especially storms that never make landfall. That’s one of the reasons for the focus on landfalling hurricanes to begin with. With the advent of satellite data, storms are highly unlikely to be missed, but even landfalls have sometimes gone unreported historically. The farther back one goes, the less is known about the extent of hurricane activity, but Pielke and Maue feel that post-1970 data is fairly comprehensive.

The chart at the top of this post is a summery of the data that Pielke and Maue have compiled. There are no obvious trends in terms of the number of storms or their strength. The 1970s were quiet while the 90s were more turbulent. The absence of trends also characterizes NOAA’s data on U.S. landfalling hurricanes since 1851, as noted by Pail Driessen. Here is Driessen on Florida hurricane history:

“Using pressure, Ian was not the fourth-strongest hurricane in Florida history but the tenth. The strongest hurricane in U.S. history moved through the Florida Keys in 1935. Among other Florida hurricanes stronger than Ian was another Florida Keys storm in 1919. This was followed by the hurricanes in 1926 in Miami, the Palm Beach/Lake Okeechobee storm in 1928, the Keys in 1948, and Donna in 1960. We do not know how strong the hurricane in 1873 was, but it destroyed Punta Rassa with a 14-foot storm surge. Punta Rassa is located at the mouth of the river leading up to Ft. Myers, where Ian made landfall.”

Neil L. Frank, veteran meteorologist and former head of the National Hurricane Center, bemoans the changed conventions for assigning names to storms in the satellite era. A typical clash of warm and cold air will often produce thunderstorms and wind, but few of these types of systems were assigned names under older conventions. They are not typical of systems that usually produce tropical cyclones, although they can. Many of those kinds of storms are named today. Right or wrong, that gives the false impression of a trend in the number of named storms. Not only is it easier to identify storms today, given the advent of satellite data, but storms are assigned names more readily, even if they don’t strictly meet the definition of a tropical cyclone. It’s a wonder that certain policy advocates get away with saying the outcome of all this is a legitimate trend!

As Frank insists, there is no evidence of a trend toward more frequent and powerful hurricanes during the last several decades, and there is no evidence of rapid intensification. More importantly, there is no evidence that climate change is leading to more hurricane activity. It’s also worth noting that today we suffer far fewer casualties from hurricanes owing to much earlier warnings, better precautions, and better construction.

Hiring Discrimination In the U.S., Canada, and Western Europe

10 Monday Oct 2022

Posted by in Discrimination

≈ Leave a comment

Tags

, , , , , , , , ,

Some people have the impression that the U.S. is uniquely bad in terms of racial, ethnic, gender, and other forms of discrimination. This misapprehension is almost as grossly in error as the belief held in some circles that the history of slavery is uniquely American, when in fact the practice has been so common historically, and throughout the world, as to be the rule rather than the exception.

This week, Alex Tabarrok shared some research I’d never seen on one kind of discriminatory behavior. In his post, “The US has Relatively Low Rates of Hiring Discrimination”, he cites the findings of a 2019 meta-study of “… 97 Field Experiments of Racial Discrimination in Hiring”. The research focused on several Western European countries, Canada, and the U.S. The experiments involved the use of “faux applicants” for actual job openings. Some studies used applications only and were randomized across different racial or ethnic cues for otherwise similar applicants. Other studies paired similar individuals of different racial or ethnic background for separate in-person interviews.

The authors found that hiring discrimination is fairly ubiquitous against non-white groups across employers in these countries. The authors were careful to note that the study did not address levels of hiring discrimination in countries outside the area of the study. They also disclaimed any implication about other forms of discrimination within the covered countries, such as bias in lending or housing.

The study’s point estimates indicated “ubiquitous hiring discrimination”, though not all the estimates were statistically significant. My apologies if the chart below is difficult to read. If so, try zooming in, clicking on it, or following the link to the study above.

Some of the largest point estimates were highly imprecise due to less coverage by individual studies. The impacted groups and severity varied across countries. Blacks suffered significant discrimination in the U.S., Canada, France, and Great Britain. For Hispanics, the only coverage was in the U. S. and sparsely in Canada. The point estimates showed discrimination in both counties, but it was (barely) significant only in the U.S. For Middle Eastern and North African (MENA) applicants, discrimination was severe in France, the Netherlands, Belgium, and Sweden. Asian applicants faced discrimination in France, Norway, Canada, and Great Britain.

Across all countries, the group suffering the least hiring discrimination was white immigrants, followed by Latin Americans / Hispanics (but only two countries were covered). Asians seemed to suffer the most discrimination, though not significantly more than Blacks (and less in the U.S. than in France, Norway, Canada, and Great Britain). Blacks and MENA applicants suffered a bit less than Asians from hiring discrimination, but again, not significantly less.

Comparing countries, the authors used U.S. hiring discrimination as a baseline, assigning a value of one. France had the most severe hiring discrimination and at a high level of significance. Sweden was next highest, but it was not significantly higher than in the U.S. Belgium, Canada, the Netherlands and Great Britain had higher point estimates of overall discrimination than the U. S., though none of those differences were significant. Employers in Norway were about as discriminatory as the U.S., and German employers were less discriminatory, though not significantly.

The upshot is that as a group, U.S. employers are generally at the low end of the spectrum in terms of discriminatory hiring. Again, the intent of this research was not to single out the selected countries. Rather, these countries were chosen because relevant studies were available. In fact, Tabarrok makes the following comment, which the authors probably wouldn’t endorse and is admittedly speculative, but I suspect it’s right:

“I would bet that discrimination rates would be much higher in Japan, China and Korea not to mention Indonesia, Iraq, Nigeria or the Congo. Understanding why discrimination is lower in Western capitalist democracies would reorient the literature in a very useful way.”

So the U.S. is not on the high-side of this set of Western countries in terms of discriminatory hiring practices. While discrimination against blacks and Hispanics in the U.S. appears to be a continuing phenomenon, overall hiring discrimination in the U.S. is, at worst, comparable to many European countries.

To anticipate one kind of response to this emphasis, the U.S. is not alone in its institutional efforts to reduce discrimination. In fact, the study’s authors say:

“A fairly similar set of antidiscrimination laws were adopted in North America and many Western European countries from the 1960s to the 1990s. In 2000, the European Union passed a series of race directives that mandated a range of antidiscrimination measures to be adopted by all member states, putting their legislative frameworks on racial discrimination on highly similar footing.”

Despite these similarities, there are a few institutional details that might have some bearing on the results. For example, France bans the recording and “formal discussion” of race and ethnicity during the hiring process. (However, photos are often included in job applications in European countries.) Does this indicate that reporting mandates and prohibiting certain questions reduce hiring discrimination? That might be suggestive, but the evidence is not as clear cut as the authors seem to believe. They cite one piece of conflicting literature on that point. Moreover, it does not explain why Great Britain had a greater (and highly significant) point estimate of discrimination against Asians, or why Canada and Norway were roughly equivalent to France on this basis. Nor does it explain why Sweden and Belgium did not differ from France significantly in terms of discrimination against MENA applicants. Or why Canada was not significantly different from France in terms of hiring discrimination against Blacks. Overall, discrimination in Sweden was not significantly less than in France. Still, at least based on the three applicant groups covered by studies of France, that country had the highest overall level of discrimination. France also had the most significant departure from the U.S., where recording the race and ethnicity of job applicants is institutionalized.

Germany had the lowest overall point estimates of hiring discrimination in the study. According to the authors, employers in German-speaking countries tend to collect a fairly thorough set of background information on job applications. This detail can actually work against discrimination in hiring. Tabarrok notes that so-called “ban the box” policies, or laws that prohibit employers from asking about an applicant’s criminal record, are known to result in greater racial disparities in hiring. The same is true of policies that threaten sanctions against the use of objective job qualifications which might have disparate impacts on “protected” groups. That’s because generalized proxies based on race are often adopted by hiring managers, consciously or subconsciously.

Discrimination in hiring based on race and ethnicity might actually be reasonable when a job entails sensitive interactions requiring high levels of trust with members of a minority community. This statement acknowledges that we do not live in a perfect world in which racial and ethnic differences are irrelevant. Still, aside from exceptions of that kind, overt hiring discrimination based on race or ethnicity is a negative social outcome. The conundrum we face is whether it is more or less negative than efforts to coerce nondiscrimination on those bases across a broad range of behaviors, most of which are nondiscriminatory to begin with, and when interventions often have perverse discriminatory effects. Policymakers and observers in the U.S. should maintain perspective. Discriminatory behavior persists in the U.S., especially against Blacks, but some of this discrimination is likely caused by prohibitions on objective tests of relevant job skills. And as the research discussed above shows, employers here appear to be a bit less discriminatory than those in most other Western democracies.

“Hard Landing” Is Often Cost of Fixing Inflationary Policy Mistakes

05 Wednesday Oct 2022

Posted by in Inflation, Monetary Policy

≈ Leave a comment

Tags

, , , , , , , , , , , , , , , , , , , , , , ,

The debate over the Federal Reserve’s policy stance has undergone an interesting but understandable shift, though I disagree with the “new” sentiment. For the better part of this year, the consensus was that the Fed waited too long and was too dovish about tightening monetary policy, and I agree. Inflation ran at rates far in excess of the Fed’s target, but the necessary correction was delayed and weak at the start. This violated the necessary symmetry of a legitimate inflation-targeting regime under which the Fed claims to operate, and it fostered demand-side pressure on prices while risking embedded expectations of higher prices. The Fed was said to be “behind the curve”.

Punch Bowl Resentment

The past few weeks have seen equity markets tank amid rising interest rates and growing fears of recession. This brought forth a chorus of panicked analysts. Bloomberg has a pretty good take on the shift. Hopes from some economists for a “soft landing” notwithstanding, no one should have imagined that tighter monetary policy would be without risk of an economic downturn. At least the Fed has committed to a more aggressive policy with respect to price stability, which is one of its key mandates. To be clear, however, it would be better if we could always avoid “hard landings”, but the best way to do that is to minimize over-stimulation by following stable policy rules.

Price Trends

Some of the new criticism of the Fed’s tightening is related to a perceived change in inflation signals, and there is obvious logic to that point of view. But have prices really peaked or started to reverse? Economist Jeremy Siegel thinks signs point to lower inflation and believes the Fed is being too aggressive. He cites a series of recent inflation indicators that have been lower in the past month. Certainly a number of commodity prices are generally lower than in the spring, but commodity indices remain well above their year-ago levels and there are new worries about the direction of oil prices, given OPEC’s decision this week to cut production.

Central trends in consumer prices show that there is a threat of inflation that may be fairly resistant to economic weakness and Fed actions, as the following chart demonstrates:

Overall CPI growth stopped accelerating after June, and it wasn’t just moderation in oil prices that held it back (and that moderation might soon reverse). Growth of the Core CPI, which excludes food and energy prices, stopped accelerating a bit earlier, but growth in the CPI and the Core CPI are still running above 8% and 6%, respectively. More worrisome is the continued upward trend in more central measures of CPI growth. Growth in the median component of the CPI continues to accelerate, as has the so-called “Trimmed CPI”, which excludes the most extreme sets of high and low growth components. The response of those central measures lagged behind the overall CPI, but it means there is still inflationary momentum in the economy. There is a substantial risk that expectations of a more permanent inflation are becoming embedded in expectations, and therefore in price and wage setting, including long-term contracts.

The Fed pays more attention to a measure of prices called the Personal Consumption Expenditures (PCE) deflator. Unlike the CPI, the PCE deflator accounts for changes in the composition of a typical “basket” of goods and services. In particular, the Fed focuses most closely on the Core PCE deflator, which excludes food and energy prices. Inflation in the PCE deflator is lower than the CPI, in large part because consumers actively substitute away from products with larger price increases. However, the recent story is similar for these two indices:

Both overall PCE inflation and Core PCE inflation stopped accelerating a few months ago, but growth in the median PCE component has continued to increase. This central measure of inflation still has upward momentum. Again, this raises the prospect that inflationary forces remain strong, and that higher and more widespread expected inflation might make the trend more difficult for the Fed to rein in.

That leaves the Fed little choice if it hopes to bring inflation back down to its target level. It’s really a only a choice of whether to do it faster or slower. One big qualification is that the Fed can’t do much about supply shortfalls, which have been a source of price pressure since the start of the rebound from the pandemic. However, demand pressures have been present since the acceleration in price growth began in earnest in early 2021. At this point, it appears that they are driving the larger part of inflation.

The following chart shows share decompositions for growth in both the “headline” PCE deflator and the Core PCE deflator. Actual inflation rates are NOT shown in these charts. Focus only on the bolder colored bars. (The lighter bars represent estimates having less precision.) Red represents “supply-side” factors contributing to changes in the PCE deflator, while blue summarizes “demand-side” factors. This division is based on a number of assumptions (methodological source at the link), but there is no question that demand has contributed strongly to price pressures. At least that gives a sense about how much of the inflation can be addressed by actions the Fed might take.

I mentioned the role of expectations in laying the groundwork for more permanent inflation. Expected inflation not only becomes embedded in pricing decisions: it also leads to accelerated buying. So expectations of inflation become a self-fulfilling prophesy that manifests on both the supply side and the demand-side. Firms are planning to raise prices in 2023 because input prices are expected to continue rising. In terms of the charts above, however, I suspect this phenomenon is likely to appear in the “ambiguous” category, as it’s not clear that the counting method can discern the impacts of expectations.

What’s a Central Bank To Do?

Has the Fed become too hawkish as inflation accelerated this year while proving to be more persistent than expected? One way to look at that question is to ask whether real interest rates are still conducive to excessive rate-sensitive demand. With PCE inflation running at 6 – 7% and Treasury yields below 4%, real returns are still negative. That’s hardly seems like a prescription for taming inflation, or “hawkish”. Rate increases, however, are not the most reliable guide to the tenor of monetary policy. As both John Cochrane and Scott Sumner point out, interest rate increases are NOT always accompanied by slower money growth or slowing inflation!

However, Cochrane has demonstrated elsewhere that it’s possible the Fed was on the right track with its earlier dovish response, and that price pressures might abate without aggressive action. I’m skeptical to say the least, and continuing fiscal profligacy won’t help in that regard.

The Policy Instrument That Matters

Ultimately, the best indicator that policy has tightened is the dramatic slowdown (and declines) in the growth of the monetary aggregates. The three charts below show five years of year-over-year growth in two monetary measures: the monetary base (bank reserves plus currency in circulation), and M2 (checking, saving, money market accounts plus currency).

Growth of these aggregates slowed sharply in 2021 after the Fed’s aggressive moves to ease liquidity during the first year of the pandemic. The monetary base and M2 growth have slowed much more in 2022 as the realization took hold that inflation was not transitory, as had been hoped. Changes in the growth of the money stock takes time to influence economic activity and inflation, but perhaps the effects have already begun, or probably will in earnest during the first half of 2023.

The Protuberant Balance Sheet

Since June, the Fed has also taken steps to reduce the size of its bloated balance sheet. In other words, it is allowing its large holdings of U.S. Treasuries and Agency Mortgage-Backed Securities to shrink. These securities were acquired during rounds of so-called quantitative easing (QE), which were a major contributor to the money growth in 2020 that left us where we are today. The securities holdings were about $8.5 trillion in May and now stand at roughly $8.2 trillion. Allowing the portfolio to run-off reduces bank reserves and liquidity. The process was accelerated in September, but there is increasing tension among analysts that this quantitative tightening will cause disruptions in financial markets and ultimately the real economy, There is no question that reducing the size of the balance sheet is contractionary, but that is another necessary step toward reducing the rate of inflation.

The Federal Spigot

The federal government is not making the Fed’s job any easier. The energy shortages now afflicting markets are largely the fault of misguided federal policy restricting supplies, with an assist from Russian aggression. Importantly, however, heavy borrowing by the U.S. Treasury continues with no end in sight. This puts even more pressure on financial markets, especially when such ongoing profligacy leaves little question that the debt won’t ever be repaid out of future budget surpluses. The only way the government’s long-term budget constraint can be preserved is if the real value of that debt is bid downward. That’s where the so-called inflation tax comes in, and however implicit, it is indeed a tax on the public.

Don’t Dismiss the Real Costs of Inflation

Inflation is a costly process, especially when it erodes real wages. It takes its greatest toll on the poor. It penalizes holders of nominal assets, like cash, savings accounts, and non-indexed debt. It creates a high degree of uncertainty in interpreting price signals, which ordinarily carry information to which resource flows respond. That means it confounds the efficient allocation of resources, costing all of us in our roles as consumers and producers. The longer it continues, the more it erodes our economy’s ability to enhance well being, not to mention the instability it creates in the political environment.

Imminent Recession?

So far there are only limited signs of a recession. Granted, real GDP declined in both the first and second quarters of this year, but many reject that standard as overly broad for calling a recession. Moreover, consumer spending held up fairly well. Employment statistics have remained solid, though we’ll get an update on those this Friday. Nevertheless, payroll gains have held up and the unemployment rate edged up to a still-low 3.7% in August.

Those are backward-looking signs, however. The financial markets have been signaling recession via the inverted yield curve, which is a pretty reliable guide. The weak stock market has taken a bite out of wealth, which is likely to mean weaker demand for goods. In addition to energy-supply shocks, the strong dollar makes many internationally-traded commodities very costly overseas, which places the global economy at risk. Moreover, consumers have run-down their savings to some extent, corporate earnings estimates have been trimmed, and the housing market has weakened considerably with higher mortgage rates. Another recent sign of weakness was a soft report on manufacturing growth in September.

Deliver the Medicine

The Fed must remain on course. At least it has pretensions of regaining credibility for its inflation targeting regime, and ultimately it must act in a symmetric way when inflation overshoots its target, and it has. It’s not clear how far the Fed will have to go to squeeze demand-side inflation down to a modest level. It should also be noted that as long as supply-side pressures remain, it might be impossible for the Fed to engineer a reduction of inflation to as low as its 2% target. Therefore, it must always bear supply factors in mind to avoid over-contraction.

As to raising the short-term interest rates the Fed controls, we can hope we’re well beyond the halfway point. Reductions in the Fed’s balance sheet will continue in an effort to tighten liquidity and to provide more long-term flexibility in conducting operations, and until bank reserves threaten to fall below the Fed’s so-called “ample reserves” criterion, which is intended to give banks the wherewithal to absorb small shocks. Signs that inflationary pressures are abating is a minimum requirement for laying off the brakes. Clear signs of recession would also lead to more gradual moves or possibly a reversal. But again, demand-side inflation is not likely to ease very much without at least a mild recession.