The Fed Stays on Hold as Uncertainty Reigns

The Federal Reserve stayed on the sidelines this week as it awaits more clarity on how the Administration’s tariffs and other policies are impacting economic growth, the labor market, and inflation. Chair Powell acknowledged some cooling in U.S. economic growth and the labor market since the last FOMC meeting in May, but “nothing troubling”. In other words, not enough cooling to coax the FOMC off the sidelines and start cutting interest rates sooner to support the economy. (For a deeper dive into the policy announcement, see Michael Gregory's Thought, “Fed Not Fed-Up with Waiting”.)

The escalating war between Israel and Iran and the possibility of U.S. intervention adds a troubling new dimension to the Fed’s calculation on inflation. Oil prices have soared on the heightened risk that crude shipments that transit the Persian Gulf could be disrupted. Iran has threatened to close the Strait of Hormuz in response to the attacks, which could cut off 20% of global oil and 30% of LNG production at least for a time.

West Texas Intermediate Crude Oil futures prices have already jumped 33% since May 5^th (Chart 1). Sizable energy price declines in March and May helped bring overall consumer and producer inflation down over the past three months. Oil prices have already undone all their improvement since mid-January and as of Thursday were sitting at their highest level since July 2024. For a time on Thursday crude oil futures were trading at $77.58 per barrel, their highest level since June 2022.

A general rule of thumb is that a sustained $10 increase in oil prices could push up headline CPI inflation between 0.2 to 0.4 percentage points over the coming year, increasing prices in transportation, utilities, food, and housing. With crude oil futures prices now up nearly $20 from its May low, oil prices could tack on another 0.4 to 0.8 percentage points to headline CPI inflation. Of course, the oil price surge may yet prove temporary if the conflict de-escalates and oil supplies are not severely disrupted. We would also see a much smaller pass-through into core measures of inflation that the Fed watches more closely. Even so, the oil price surge adds to the inflation uncertainty from the tariffs and could seriously undermine the improvement we have seen in inflation expectations in May.

Chair Powell left the door open for rate cuts later this year, with the FOMC median forecast still looking for 2 quarter-point cuts (Chart 2). However, there is a growing divergence of opinion on next steps. Seven FOMC participants are already in the “no rate cut” camp for 2025. The conflict in the Middle East and surge in crude oil prices could lead more participants into the camp in the months ahead, delaying the next Fed rate cut beyond September.

The FOMC marked down their median projection for GDP growth in 2025 and 2026, slightly raised their unemployment rate call, and increased their inflation forecasts. Even so, we don’t think the Fed’s forecasts are yet stagflationary enough. Our baseline forecast shows a somewhat slower GDP growth rate of 1.0% Q4/Q4, higher unemployment rate of 4.7% by Q4, and a higher core PCE inflation rate of 3.3% Q4/Q4.

The U.S. economy has been performing very close to our expectations since the April 2^nd Liberation Day shock and May 12^th tariff truce. The latest “hard” economic indicators for May and June reveal more growth deterioration even as “softer” sentiment measures have stabilized or improved. May retail sales and industrial production both declined more than the consensus expected, while initial and continuing jobless claims have jumped to multi-year highs (Chart 3). This week, housing starts, Empire State and NAHB Housing Market Index for June all declined, missing economist forecasts. No recession visible on the horizon, but uncertainty continues to reign.

Fed Not Fed-Up with Waiting

The FOMC left policy rates unchanged on June 18, keeping the fed funds target range at 4.25%-to-4.50% for the fourth consecutive confab. The half-year hold reflects the rocketing uncertainty over the Administration’s erratically evolving economic policies and their consequences for the economy (Chart 1). And the FOMC signaled that it was in no hurry to halt the hold.

The policy statement said: “Uncertainty about the economic outlook has diminished but remains elevated. The Committee is attentive to the risks to both sides of its dual mandate.” This risk assessment was less alarming than the last one (on May 7), probably reflecting the eye-watering 125%-to-145% U.S.-China bilateral tariffs being chopped to 10%-to-30%. But the key words in the first sentence were that uncertainty “remains elevated”.

For example, from a trade policy perspective, steel and aluminum duties were doubled to 50%, ‘reciprocal’ tariffs are slated to return next month, and the appeals court ruling on the legality of tariffs justified by the International Emergency Economic Powers Act (IEEPA) is not due until August. With the de-escalation of the U.S.-China tariff war, the stagflation risk siren sounded last meeting has been turned down (the phrase, “the risks of higher unemployment and higher inflation have risen”, was removed), but the siren has not been turned off.

The median forecasts in the Summary of Economic Projections (SEP) showed, again, slower growth and faster inflation. For 2025 (Q4/Q4), real GDP growth was lowered by 0.3 ppts to 1.4% and both total and core PCE inflation were lifted by 0.3 ppts to 3.0% and 3.1%, respectively. For 2026, growth was decreased by 0.2 ppts to 1.6% and both total and core inflation were increased by 0.2 ppts to 2.4%. (In the prior SEP, for example, 2025 growth was sliced by 0.4 ppts with total and core inflation lifted by 0.2 and 0.3 ppts, respectively.)

In the ‘dot plot’, the median projection for this year stayed at two 25 bp rate cuts (to 3.875%), although one more participant joined the ‘one-or-none’ club. It now boasts nine members with just one more needed to turn the median dial. (Indeed, seven are now in the ‘none’ camp versus four before.) For next year, the median call is for only one cut—it was two before. And the missing move is not being made up for later. The stagflation-skewed forecast alterations are prompting more projected Fed policy caution.

Elsewhere, the statement still referred to “solid” economic and labor market performance. And Chair Powell acknowledged in the presser that, although this performance had slowed, recent labor developments were still not “troubling” while economic growth was still “decent”. These assessments are affording the Fed time to wait and see how tariffs are influencing the inflation process. Chair Powell said that, owing to factors such as pre-tariff stockpiling, the impact of tariffs on selling prices is being delayed. The late-winter and spring tariff increases are expected to influence prices as the summer unfolds. And the Fed is prepared to wait it out.

Bottom Line: The Fed is weighing the risks of slower growth against those of faster inflation. If the former is significantly heavier, the Fed will likely ease. But if the latter is meaningfully heavier, the Fed will probably remain on hold even longer (recall that policy is still on the restrictive side). By summer’s end, we should be starting to see a lasting data tilt to slower growth, setting the stage for a rate cut or two, pending what happens on the policy front (both for trade and fiscal policy) along with the geopolitical front (via any lasting impact on oil prices).

Meanwhile, Chair Powell will have another opportunity to discuss all these issues, with the semi-annual congressional testimony on the Fed’s Monetary Policy Report slated for next week. The report was released on Friday, with nothing majorly standing out.

LAIoffs

This was a bad week for anyone who thought AI wouldn’t have a transformational effect on the workforce. Three of the world’s largest companies weighed in on the debate. Amazon’s CEO said the company plans to downsize its staff in the coming years as AI boosts efficiency and displaces some jobs. Microsoft announced plans to lay off thousands more workers starting in July, partly to free up funds for further AI investments. This comes after cutting about 3% of its workforce in May—cuts led by software developers now competing with AI-coding bots. And, Nvidia is planning to use humanoid robots at a new factory in Texas to help produce AI servers. So, AI-driven robots building AI-systems gear. Nvidia’s CEO believes humanoid robots will be deployed widely across factories within five years.

Simply put, AI will eventually allow fewer workers to do more work in a wide range of industries. Shopify recently instructed its managers to explore AI solutions before hiring new staff. IBM is using AI to replace human-resource employees. Bank of America’s CEO said banks will need fewer people as AI is used to process payments and manage customer finances. This aligns with a recent World Economic Forum survey that found about 41% of global companies plan to reduce staff due to AI. Compared with other trends, AI (and information processing) was cited by the most companies (86%) as likely to transform their business.

There may be hope, however. Outside of the information technology sector, generative AI is still too new, and in many cases too unreliable, to replace a broad range of tasks. It still requires human oversight. Accordingly, the World Economic Forum survey also found that 77% of companies plan to train existing staff to work alongside AI, using the technology to complement rather than replace workers. Human and AI collaboration may prove to be the winning formula, with each augmenting the other’s weaknesses. For this reason, Amazon’s CEO said he doesn’t expect mass layoffs, though AI will change how tasks are done and push some workers into other roles within the company.

So far, AI is everywhere—except in the employment data (to paraphrase Robert Solow’s famous quote on productivity). No company in Challenger’s survey has cited AI as a reason for layoffs since last summer. But this could change, slowly at first, and then suddenly. Moreover, just as trade-war uncertainty has frozen some hiring plans, some companies may hesitate to hire without first testing whether existing staff can use AI (or humanoids) to handle the workload.

We’re in the early innings—still unsure whether AI will produce more home runs or strikeouts for workers. The optimist in me leans toward the former, supported by centuries of data showing that transformational technologies, over time, lead to more jobs overall. But AI feels different, and I increasingly worry about the downside for a broad range of workers, which, unlike before, now includes the highly skilled.

Is Reshoring Worth It?

The U.S. government’s reshoring drive may or may not pay off, but the net rewards could prove limited.

The White House’s list of reasons for tariffs is long, but bringing production home is near the top. Hefty levies have already been slapped on autos, steel and aluminum, and more are planned for medicines, microchips and aircraft, among others. The government also wants to reinstate full expensing of new plants to incentivize goods production. Other countries also want to develop critical industries, such as electric vehicles and batteries. Sidestepping whether tariffs (the stick) are the best way to incent reshoring—tax credits (the carrot) might be better—this note addresses whether the goal is even worth pursuing.

First, some background. U.S. manufacturers have mostly spun their wheels in the past decade (Chart 1). Apart from computer makers, nearly all industries have struggled to expand. Since peaking at over one-quarter in 1953, manufacturing’s share of the economy has shrunk to just one-tenth today. Business services eclipsed it more than a decade ago (Chart 2). Manufacturing’s shrinking role in the economy is not just confined to the U.S., as even countries with goods trade surpluses (China, Germany, and Japan) have faced a downhill battle.

At under 13 million, the U.S. factory workforce is no larger than six years ago and down from 20 million in 1979. Since peaking at 39% in World War II, manufacturing’s share of total payrolls has fallen to a record-low 8% as companies increased automation and sought out lower-cost overseas suppliers. Despite rapid output growth, even computer makers could only keep their workforce stable since 2010. Meanwhile, the private-service sector employs a record 114 million people today, or 72% of the total workforce.

With this preamble in mind, here are eight common arguments for reshoring manufacturing:

1) Enhance national security: The theory of comparative advantage says that a country should play to its strengths, doing what it does best and trading for other stuff. However, sacrificing some efficiency for national security can pay off in some cases. It would be unwise to rely on another country for critical military hardware, for example, especially if it turns hostile. The same might be said for an unhealthy reliance on foreign-made AI microchips that will likely run much of the economy someday. In both cases, developing self-sufficiency in producing certain goods, even at a higher cost than importing, might make sense.

2) Reinforce supply chains: An over-reliance on critical supplies and materials from other countries can expose domestic businesses to supply disruptions, whether from shipping delays, geopolitical conflicts, pandemics, or trade wars. U.S. automakers recently faced this threat after China (temporarily) curbed shipments of rare earth minerals. Producing domestically can remove some global impediments, adding resiliency to supply chains while also speeding up product flow.

3) Nurture local economies: A factory that hires thousands of workers can anchor the local economy by nurturing an ecosystem of suppliers, restaurants, and service providers around the plant. Of course, this can also be a curse should the factory shut down.

4) Drive construction activity: Building a large factory requires workers, materials, and machinery. Between 2021 and 2024, the funding and incentives provided by the Infrastructure Investment and Jobs Act, the Inflation Reduction Act, and the CHIPS and Science Act led to a tripling of spending on manufacturing facilities. This, in turn, drove a 15% increase in nonresidential construction jobs, nearly double the overall rate.

5) Reduce trade imbalances: Although the U.S. trade surplus in services reached a record high in 2024, it has grown only modestly in the past six years and is only one-forth the size of the goods deficit. While foreign demand for U.S. business, information and financial services, as well as tourism, will grow over time, it is unlikely to shrink the overall trade gap much. Goods trade will need to do the heavy lifting (reducing the budget deficit would also help).

6) Factories pay well (though their edge is fading): Going back to 2019, average hourly earnings in manufacturing kept pace with the private sector, but a modest gap has since opened (Chart 3). While factories still pay more than retailers or hospitality providers, the difference is narrowing. Other industries, including information, finance, mining, construction, and business services, pay considerably more. Today, more Americans work in higher-paying skilled trades (e.g., electricians, carpenters) than on the factory floor. Still, modern factories also rely on mechanical engineers, repair technicians, and computer specialists, all of whom earn above-average wages.

7) Making factories productive again: From the mid-1990s to the financial crisis, American manufacturing saw a productivity renaissance driven by new IT systems and automation (Chart 4). But the sector has stalled since 2010 while others have flourished (this partly explains the slower wage growth). Still, AI-driven automation could reignite productivity, making the sector an important driver of profits, wealth and income again.

8) Factory jobs could be less exposed to AI displacement: Humanoids, or robots with human-like abilities due to AI, will likely multiply in the decade ahead, displacing many jobs even as they create new ones. Yet, workers in knowledge-based services could be even more at risk than those in factories, which will still require highly-skilled technicians for maintenance and troubleshooting. That said, manufacturing is also highly capital-intensive, meaning that reshoring may not generate many new jobs.

Bottom Line: Time will tell whether reshoring pays off. Much like agriculture (which went through a similar debate at one time), reversing a near century-long decline in manufacturing’s role in the economy could prove daunting. Growing an industry, such as shipbuilding or semiconductors, is costly and can take many years. Given immigration curbs, there is also the question of where new workers will be found. Moreover, tastes have changed. Consumers are spending twice as much on services than goods today, compared with an equal amount in 1970. Shoppers also have a price point in which lower-cost imports (even with a tariff) become more appealing. All in, the net rewards from reshoring may be limited.