Viewpoints: 2026

To my thinking, the surefire marker for the adoption of artificially intelligent (AI) “agents” is a decline in labor’s share of income. Here’s the logic in terms of production theory.

When I learned Solow growth accounting, the constancy of labor’s share of income was an accepted “stylized” economic fact. Cobb and Douglas demonstrated that if inputs are paid their marginal product in competitive factor markets, the constancy of labor’s share was consistent with a unitary elasticity of substitution between capital and labor. The logic was easy to grasp. A decline in the cost of capital relative to the wage rate encouraged a proportionate increase in ratio of capital to labor. Therefore, payments made to capital, relative to payments made to labor, remained unchanged.

AI agents are near-perfect substitutes for workers in certain occupations. Hence, the introduction of AI agents raises the overall elasticity of substitution between capital and labor relative to the traditional unitary value. At the same time, the cost of AI agents, reflected in the price per “token” in LLMs, is falling rapidly. Given the high elasticity of substitution between AI agents and certain workers, the rapid decline in the relative cost of AI agents encourages an increase in the ratio of capital to labor that is more than proportionate to the decline in the relative cost of capital. Consequently, capital’s share of income rises, while labor’s share declines.

After remaining fairly constant for decades following WWII, labor’s share fell unevenly between 2000 and 2022 by approximately 8 percentage points, from 69% to 61%. A plausible explanation of that decline was the initial emergence of agentic capital as the dot-com era evolved. Today there’s a widespread expectation that surging investment in AI will perpetuate this decline, perhaps dramatically. Maybe so, but it hasn’t happened yet.

The chart shows the recent quarterly history (through 2026 Q1) of labor’s share of income in the private nonfarm business sector, adjusted to exclude sectoral taxes on production and imports from the denominator. This exclusion, required by theory, allows that revenues from “indirect” taxes are not received by producers, and therefore not allocable to either capital or labor. When calculating “labor share” for its report on productivity and costs, the Bureau of Labor Statistics doesn’t make this adjustment. However, given the recent surge in customs duties, almost all of which are levied on private nonfarm business, the correction is important.

Perhaps aggregation conceals nascent effects in vulnerable sectors. Perhaps it will take more time for the ongoing “AI buildout” to affect the distribution of income noticeably in any sector. Perhaps the data will be revised. Be that as it may, more three and one half years after the introduction of ChatGPT in November of 2022, there has been no decline in labor’s share of income when correctly measured.

Europe entered the latest energy shock from a weaker position than it occupied before the 2025 global tariff shock. Unlike much of Asia, its exporters have yet to regain momentum and economic growth has slowed. First-quarter GDP data underline the divergence. Euro-area output expanded by only 0.8% year on year, weaker than at the end of last year and below the pace recorded a year earlier. By contrast, growth accelerated in the United States, Korea, Taiwan and China, while Japan broadly held steady (Figure1). Europe therefore enters the latest period of geopolitical uncertainty with less economic momentum than many of its major competitors, leaving businesses and policymakers with less room to absorb further shocks.

The Federal Reserve instituted the payment of interest on reserves on October 15, 2008, during the Great Financial Crisis. The motivating factor for this was that the federal funds rate was trading below the FOMC’s target rate. Banks had greatly increased their borrowing from the Federal Reserve. This created excess reserves (reserves in excess of then required reserves) in the banking system. These zero-yielding excess reserves put downward pressure on the federal funds rate. In order to induce banks to hold excess reserves and, thus, prevent the federal funds rate from falling below the FOMC’s target-rate level, the Federal Reserve began paying interest on reserves. By December 17, 2008, the FOMC had reduced the lower-limit of its target federal funds to zero. So, the Federal Reserve could have ceased paying interest on reserves because the FOMC’s target-level of federal funds rate was at zero, a level at which the actual federal funds rate could not fall below. Yet, the Federal Reserve persisted in paying interest of reserves.

The FOMC began raising its FOMC target-level of the federal funds rate above effective zero in December 2015. Yet the Federal Reserve persisted in paying interest on reserves, which it does today with a FOMC target level of the federal funds rate at 3.625%. In February 2009, the FOMC engaged in the first of a series of Quantitative Easing (QE) operations whereby the Federal Reserve added large quantities of securities to its outright holdings, culminating in the largest QE operation in 2020 during the Covid pandemic. On April 1, 2020, the Federal Reserve eliminated reserve requirements on banks. With the massive amounts of reserves created by the Federal Reserve since 2009 and with the elimination of reserve requirements in 2020, how is the FOMC able to maintain the federal funds rate above zero? It does so by the artificially-created demand for reserves resulting from the payment of interest on reserves by the Federal Reserve.

If Chairman Warsh wishes to reduce the size of the Federal Reserve’s balance sheet, I suggest that the Federal Reserve cease paying interest on reserves. Banks’ demand for reserves would fall significantly (but not to zero). This would enable the Federal Reserve to offload large quantities of securities in order to drain off the “excess” and unwanted supply of reserves. If the FOMC insists on continuing to conduct monetary policy through the federal funds rate, it could announce that it would lend reserves via repurchase agreements at the federal funds target level plus X basis points or it would borrow reserves via reverse repurchase agreements at the federal funds rate minus X basis points.

The Federal Reserve owns $6.4 trillion of securities outright of which $4.5 trillion are US Treasury securities. So, the Federal Reserve owns about 14-1/2% of total marketable Treasury debt outstanding. If the Federal Reserve is concerned about the market “digesting” $6.4 trillion of debt all at once, it could phase down the amount of reserves on which it would pay reserves over time. But one way or another, there is no reason why the Federal Reserve should continue to pay interest on bank reserves.

Movements in the Federal Reserve Bank of Philadelphia’s state coincident indexes in May were again, generally, fairly modest, but with a somewhat wider range than we’ve recently seen. In the one-month changes, none had increases as high as 1 percent, though seven had gains above .5 percent. On the other side, five states had declines, with Kentucky, Hawaii, and Alabama’s fairly noticeable (more than .2 percent). Over the three months since February, thirteen states had increases of 1 percent or higher, with West Virginia’s 2.70 percent far and away on top. Four states saw declines, with Alabama’s -.33 percent being the largest. Over the last twelve months, Ohio, Nevada, Idaho, California, and North Dakota clocked increases above 3 percent. Five states were down, with West Virginia’s 1.55 percent loss substantially larger than any other state’s.

The independently estimated national estimates of growth over the last three and twelve months were, respectively, .67 and 1.91 percent. These seem to be consistent with the state figures.

State real GDP growth rates in 2026:Q1 ranged from -1.6% in South Dakota to 4.5% in Washington. Declines in farm output held back South Dakota and other states in the Great Plains. Information—presumably connected to AI—boosted Washington. In general, states in the West and Southeast outperformed those in other regions.

Personal income growth rates ranged from -23.9% in Hawaii to North Dakota’s 22.4%. Those two states reversed their positions from 2025:Q4 as special factors at work reversed (in Hawaii, transfer payments, in North Dakota, net earnings. Unlike GDP, personal income growth was strongest in the Great Plains.

*The dramatic spike in oil prices generated 3 months of outsized increases in CPI and PCE inflation that pushed up their yr/yr measures.

*The Fed has kept rates on hold and did not accommodate the negative supply shock, core inflation measures have not increased much, and inflationary expectations have remained anchored.

*Oil prices have fallen sharply, and if current prices stick close to current levels (around or below $75/barrel)—which is uncertain as inventories need rebuilding--continued rapid price declines in gasoline and other energy may result in several months of declines in the CPI and PCE Price Index.

*This would ease some price pressures on consumers and let the Fed breath more easily.

In a note in mid-April, I described how the spike in oil prices from $65/barrel to $95/bbl would temporarily boost the monthly inflation data for about three months as retail prices adjusted to the higher oil prices. I emphasized the importance of the Fed not accommodating the negative supply shock, which would keep its impacts temporary, limit the pass through of the higher oil prices to the prices of nonenergy goods and services, and constrain inflationary expectations. I noted that if oil prices remained around $95/bbl, after several months of outsized increases, the monthly inflation data would revert to their prior increases, while their yr/yr measures would rise to absorb the temporary monthly spikes.

I continued: “However, if oil prices fall, subsequent months' CPI and PCE Price Index data would possibly decline, and the temporary months of deflation would reduce the general price level from its oil price-driven peak.” That process is now beginning to unfold. Yippee: a positive supply shock that involves a partial reversal of the negative supply shock imposed by the conflict in the Middle East and relief to consumers and businesses.

Here’s the situation: in the prior 12 months through February 2026, CPI inflation was 2.4%, averaging a 0.2% increase per month, while PCE inflation was 2.9%, averaging a 0.3% rise per month. The oil price spike pushed up CPI inflation 0.9%, 0.6% and 0.5% in March, April and May, lifting its yr/yr inflation to 4.2% in May, while PCE inflation rose 0.7% and 0.4% in March and April, lifting its inflation to 3.8% in April (PCE inflation for May will be reported tomorrow). See Chart 1. During these months, both core CPI inflation and PCE inflation rose a bit (CPI: 0.2%, 0.4% and 0.2%; PCE: 0.3% and 0.2%), but the details of the CPI indicate that the pass through of the oil price spike was fairly limited to specific categories, including energy services (electricity and utilities) and airline fares. Most categories in the CPI showed little effect of the higher oil prices.

During the June 17 press conference after the FOMC meeting, the new Fed Chairman, Warsh, announced the establishment of five task groups. These groups will address various topics, such as Fed communications, balance sheet policy, inflation framework, employment and productivity, and data sources that assist policymakers. The proposed changes to policymakers' practices could significantly affect how the Fed plans to achieve its objectives, compelling Wall Street to adapt to a new approach.

Fed Communication: Policymakers usually prioritize items based on their significance, which is why Fed Warsh considers Fed communication as the foremost concern. Fed Warsh has publicly stated that the "Fed" communicates excessively and too frequently. A notable shift in Fed communication was evident during his initial FOMC meeting. The press release was short, with no forward guidance, and Fed Warsh did not engage in offering forecasts for growth, unemployment, inflation, and policy rates. It wouldn't be surprising if the Fed decides to drop the "dot plot" and "forward guidance" in upcoming meetings. Wall Street might protest, but how many companies receive a "roadmap" every six weeks? Reducing risk and speculation in the financial markets is a good thing.

Fed's Balance Sheet Policy: Fed Warsh has long argued that the Fed's balance sheet is too big. It is unclear if Mr. Warsh would change the size, composition, or both, but he wants to reduce the Fed's footprint. At the June press conference, when asked if he thought Fed policy was restrictive, he argued if you look at what is happening in the financial markets, it is hard to say policy is restrictive. Financial markets run on liquidity, and Fed Warsh thinks the Fed balance sheet is providing too much liquidity.

Existing Data Sources: Fed Warsh attempts to imitate former Fed Chairman Greenspan in various aspects, yet unlike Greenspan, who was a "data junkie," Warsh does not share this trait, particularly regarding economic data. Fed Greenspan spent a lifetime studying microeconomic data and was a big fan of survey data from the purchasing managers and others. However, the most reliable and hardest data originates from government statistical agencies, including the Census Bureau, Bureau of Labor Statistics, and Bureau of Economic Analysis. If Mr. Warsh wants to get a more accurate account of the economy, he would be wise to ask Congress to increase funding for the statistical agencies.

Productivity and Jobs: Mr. Warsh is of the opinion that "AI" will have a positive impact on output, employment, and productivity, although the extent and timing of these benefits are unpredictable. If Fed Warsh takes a similar approach to Mr. Greenspan, he will allow the economy and financial markets to guide him.

Inflation Frameworks: Mr. Warsh has argued that "trimmed inflation" measures offer a better guide of underlying inflation versus widely used "core measures". Yet, if Mr.Warsh is serious about revisiting inflation frameworks, then there should be a serious discussion about inflation measurement. The Fed's price target, the PCE deflator, is not a direct measure of inflation, as nearly a third of it comes from non-consumer, non-market prices. The CPI has its flaws, and both measures include an implied rent series for homeowners. Price statistics need to be relevant, objective, and reflective of people's actual experiences. Currently, price measures do not meet these standards.

As Mr. Warsh stated, a change in leadership is a "timely opportunity to review current practices" and review what still works and what should be changed. Wall Street must adapt as the operation of monetary policy could change significantly under Fed Warsh's leadership.

State labor markets in May were again somewhat firmer. Two states saw statistically significant increases in payrolls from April: North Carolina reports a 17,400 increase (.3 percent), and West Virginia’s 9,700 increase was a whopping 1.4 percent. No state had a statistically significant decline, and only a few reported insignificant drops.

Seven states reported statistically significant drops in their unemployment rates in April, though none was larger than .2 percentage point. Alabama reported a .2 percentage point increase. Rates at or above 5.0 percent were in DC, California, Nevada, Washington, Delaware, and Illinois, with DC’s 6.1 percent the highest. Hawaii, North Dakota, South Dakota, and Vermont had unemployment rates under 3.0 percent, while South Dakota’s 2.1 percent was the lowest in the nation.

Puerto Rico's unemployment rate was unchanged at 5.6 percent and the island’s job count rose 1,500.