Government shutdown

The U.S. government shutdown began on October 1 and is now two weeks old. Our last #MacroMemo addressed some of the broad parameters surrounding such shutdowns. Only non-essential work is halted, with the implication that approximately 750,000 out of the federal government’s nearly three million workers have been furloughed.

Note that no one, including the working essential staff, is paid during the shutdown. All, including the furloughed workers, will receive back pay after the shutdown is over. Non-essential government contractors are an exception.

The initial thinking was that the prospect of missing an October 15 military pay date might provide sufficient motivation to secure a deal by then. But that date now looms and market expectations are mounting that the shutdown will extend not just for longer, but for 30+ days. Polymarket is now assigning a 69% chance of this.

At the heart of the budget dispute is whether to restore health care spending (Democrats) or not (Republicans). The Republicans also seek some budget cuts.

The White House is pressuring Democrats by pausing funding on transportation and energy products in Democratic-majority states, and threatening to permanently fire federal workers if a deal is not struck.

The Republican Party seemingly has a measure of tolerance for the damage associated with a shutdown because it fundamentally aspires to shrink the size of the government, and this is a (broad, abrupt and temporary) version of that objective.

The Democratic Party also appears willing to tolerate a shutdown given the importance it ascribes to restoring health care funding. This also appears to be the only way the party can exert even a modicum of influence in a Washington where power has become unusually concentrated in the White House.

This is not to say that the shutdown will therefore go on forever. As pay cheques are missed and public frustration mounts, the pressure on politicians to secure a deal mounts. A few of the more moderate Republican Senators are reported to be working on proposals that bridge the gap. Senate Democrats voted with the Republicans on October 9 to pass a defence policy bill, showing some ability to compromise.

The most likely scenario at this point is that both parties accede to a short-term extension lasting several weeks that would then buy politicians more time to hammer out an enduring deal. That would abruptly end the government shutdown but would present the risk of another shutdown later in the year if a deal isn’t struck.

In the meantime, economic damage is mounting. The 2018-2019 shutdown rendered the economy about 0.6% smaller than it would otherwise have been for the 35-day duration of that shutdown. That’s not a ton, especially given that economic output in the other 11 months of the year was largely unaffected (and was even a bit stronger than otherwise due to some catch-up growth). But neither is it a trivial effect.

The last shutdown only affected five of the federal government’s 12 appropriation categories, whereas this one affects all 12. Thus the damage this time should be somewhat larger. In line with the fact that roughly twice as many government workers have been furloughed this time, one might estimate that the economic damage should be about twice as large as the last time – between a 1.0% and 1.25% hit to U.S. economic output during the affected time period.

This means that if the shutdown were to last for a month, annualized gross domestic product (GDP) growth in that quarter would be about 1.5 percentage points lower, and annual GDP growth for 2025 would be about 0.1 percentage point lower.

One side-effect of the shutdown is that economic data is not being released. Most acutely, the payrolls report for September was not published on October 3. Fortunately, we have ways of estimating what might have happened:

• The privately maintained ADP employment survey for September calculated that 32,000 jobs were lost during the month.

• The weekly initial jobless claims number for the U.S. was not published due to the shutdown, but we were able to calculate it ourselves from state-level data. It shows little change in the figure from earlier weeks (see next chart).

• The Chicago Federal Reserve published an estimate of what the U.S. unemployment rate would have been if the figure had been released, and the 4.34% number is a hair above the official 4.3% number for August.

All told, the U.S. job market probably softened slightly further in September.

U.S. jobless claims remain low

RBC GAM estimates, as of the week ending 09/27/2025, based on latest state level data available. Actual claims data as of the week ending 09/20/2025. Sources: U.S. Department of Labor, Haver Analytics, RBC GAM

The next major data release is the U.S. Consumer Price Index (CPI) print for September, which would normally be published on October 15. This has now been delayed until October 24 and will be published on that date even if the shutdown persists. Some non-essential workers have been recalled to tabulate the number because it is so important, not just for economists and central bankers trying to gauge the temperature of the economy, but also for pension plans that are indexed to inflation and to determine inflation-linked bond payouts.

The U.S. Federal Reserve (Fed) has its next policy decision on October 29. The Fed has not been affected by the shutdown as it is self-funding, but the quality of its decision-making obviously suffers when economic data is missing. We believe the likelihood remains quite high that the Fed delivers a second consecutive 25-basis-point rate cut on that date despite the ambiguity. This is for a few reasons.

The Fed’s dot plots had indicated further cuts for the remainder of 2025 before the shutdown began, and the recently released Fed minutes further confirm that. The non-conventional economic data such as the ADP survey argue that the economy remains underwhelming, supporting a cut. The shutdown itself also does additional economic damage, and while that is only temporary, it could have negative spillover effects on risk-taking in other parts of the economy, doing more enduring damage.

Unsurprisingly, news-based economic sentiment has taken a visible turn for the worse (see next chart).

Daily news sentiment in the U.S. has declined

As of 10/05/2025. Sources: Federal Reserve Bank of San Francisco, S&P Global, Macrobond, RBC GAM

Tariff musings

Renewed China tensions

Tensions between the U.S. and China increased substantially in recent days, with the U.S. threatening an additional 100% tariff on Chinese goods starting November 1. This is after China announced a plan to expand its rare earth export restrictions starting November 8 to include additional tech-oriented inputs including artificial diamonds, lithium battery materials and graphite anode materials.

The U.S. also began collecting port fees from Chinese vessels entering the U.S. on October 14, and China has announced it will impose its own port fees on U.S.-connected vessels.

All of this is happening with less than a month until the November 10 deadline for the two countries to strike a trade pact. After that date, tariffs would rise further.

Financial markets are expressing concern about this antagonism between the world’s two economic superpowers. However, President Trump said over the weekend that “it will all be fine,” and Treasury Secretary Bessent announced a plan to meet with his Chinese counterpart. This prompted a partial rebound (see next chart).

Stock market tumbled over China concerns

As of 10/13/2025. Sources: S&P Global, Macrobond, RBC GAM

For our part, we do not expect a large further leap in U.S.-China tariffs, but neither do we anticipate a deal that takes tariffs down to the relatively moderate 15-20% levels achieved by most other countries.

New tariffs

The next wave of U.S. sector-oriented tariffs are arriving.

Forestry-related tariffs were implemented on October 14. These are a 10% rate on softwood lumber and timber imports, plus a 25% tariff on kitchen cabinets, bathroom vanities and upholstered furniture.

Heavy-duty truck tariffs have been postponed from October 1 to November 1, at which point a 25% rate will apply. The top import sources for such vehicles are Mexico, Canada, Japan and then Germany.

Pharmaceutical tariffs were meant to have begun on October 1, but no executive order or proclamation has been published, and so customs are not thought to be collecting them yet. The tariffs are meant to apply at a hefty 100% rate, but with rather large exemptions. Generic tariffs are expected to be excluded, and patented pharmaceutical products are exempted if the company is building a U.S. manufacturing facility. In practice, many pharmaceutical giants are doing precisely that, so the tariff – when eventually implemented – may prove quite porous.

On the net, forestry-related tariffs barely budge the overall average U.S. tariff rate higher, up by less than a tenth of a percentage point to 17.4% (see next chart). If the truck and pharmaceutical tariffs land, the tariffs could increase by up to another percentage point or two.

Average U.S. tariff rate rises slightly

Effective tariff rates estimated based on tariffs in effect at the specified date and up to 10/14/2025; threatened rates not included. Excludes the de minimis effect – suspension of de minimis exemption for China and Hong Kong in May 2025 and effective 08/29/2025 for all other countries. Expected tariff rate assumes instantaneous and complete implementation, i.e. does not account for shipping delays, implementation lags, etc. Sources: Evercore ISI Tariff Tracker, International Monetary Fund (IMF), Macrobond, RBC GAM

Effective tariff rates estimated based on tariffs in effect at the specified date and up to 10/14/2025; threatened rates not included. Excludes the de minimis effect – suspension of de minimis exemption for China and Hong Kong in May 2025 and effective 08/29/2025 for all other countries. Expected tariff rate assumes instantaneous and complete implementation, i.e. does not account for shipping delays, implementation lags, etc. Sources: Evercore ISI Tariff Tracker, International Monetary Fund (IMF), Macrobond, RBC GAM

Tariff price damage

There are new estimates of the effect that tariffs are having on U.S. consumer prices. The Harvard Business School Pricing Lab Tariff Tracker shows that, as theory would suggest, the prices of imported goods have increased distinctly more quickly than the prices of domestic American goods (see next chart).

Prices of both imported and domestic goods have risen in U.S. since February

As of 09/21/2025. Daily unweighted price indices based on online data for goods sold by four major U.S. retailers. Sources: HBS Pricing Lab Tariff Tracker, Cavallo, Llamas & Vazquez (2025), micro data from PriceStats – State Street, RBC GAM

The price change gap is only slightly more than +1% in favour of the imported goods, which is smaller than theory would dictate. A 20% average tariff should increase imported consumer goods in the U.S. by about 8%. Then again, we expect higher prices to continue passing through for some time to come.

It is also curious that the bulk of the existing price gap opened up between March and May, even though the bulk of the tariffs have been deployed since then. Perhaps companies were anticipating the higher prices in advance? More likely, there is further passthrough to come.

Tariff strategy pivoting?

Is the American tariff strategy undergoing a subtle pivot? Possibly, in a few ways.

First, it is notable that whereas the initial Trump tariff push focused on country-wide blanket tariffs – 10% on most things from the U.K., 15% on Japan, and so on – sector tariffs have since taken centre stage, with forestry, trucks and pharmaceutical products just the latest in a long list of targets.

This could simply be because the White House happens to have already implemented the bulk of its country tariffs, whereas the sector plan is more complicated and still playing out.

Still, with IEEPA (blanket-type) tariffs subject to intensifying legal scrutiny, it could be that the White House is pre-emptively pivoting toward tariffs that operate on a somewhat more solid legal footing. For context, Polymarket betting now points to just a 42% chance that the IEEPA tariffs are allowed to persist. We’ve written before about how the tariff toolkit is probably big enough to allow the White House to accomplish what it would like to over the medium run using other tools, but sector tariffs may be proving more attractive in the short run.

Another question about U.S. tariff policy is whether the focus might be expanding from goods into services. President Trump has talked of a 100% tariff on foreign-made films. It isn’t at all clear how this would be collected since services don’t physically cross the border or flow past customs officials. So far, nothing has come of it. But it wouldn’t be impossible to demand some sort of tax for foreign-made films.

This is concerning as it could open up a large swath of service industries to tariffs. The service side of the economy is far larger than the goods side, including not just the arts, but also financial services, accounting, law, consulting, information technology, education and healthcare. It is hard to conceive how some of these sectors could be tariffed, but perhaps foreign-based firms operating in the U.S. could instead be disadvantaged in some way.

-EL

Tariffs compress margins

When September’s shutdown-delayed CPI report is finally released, focus will continue to be on the extent to which tariff costs are being passed through to consumer goods prices.

There was mounting evidence of that in August as prices for many import-intensive products rose at an unusually fast rate. Even vehicle prices, which previously showed surprisingly little impact from tariffs, started to heat up. Some of the higher-frequency alternative inflation data we follow argues prices indeed continued to rise in September.

But it is nevertheless true also that there has been more tariff burden-sharing throughout the supply chain than had been anticipated. Studies of past tariff episodes indicate that most of the burden is usually borne by domestic consumers, but this time the Yale Budget Lab estimates “just” a 70% pass-through to consumers through June.

Tariff costs can alternatively be offset by exchange rate adjustments (although it hasn’t been the case this time around). Or, tariffs can be absorbed by foreign exporters or domestic manufacturers, wholesalers, and retailers – as opposed to being passed onto end consumers.

There is evidence of foreign firms absorbing a fraction of tariff costs. The price of U.S. imports from China (measured pre-tariff) are down 3% year-over-year – just a small share of the 38% effective tariff rate on China. Some of that decline may even be for other reasons, reflecting China’s general producer price deflation.

The decline in foreign pricing becomes clearer at the product level (and abstracting away from China). There is some evidence of foreign discounting in categories subject to particularly large tariff hikes, like steel and aluminum and toys, games and sporting equipment:

Foreign producers are absorbing some tariffs: U.S. import prices falling

As of 10/02/2025. Relative import price change is the change in U.S. import prices relative to export prices of the same product. Sources: U.S. Census Bureau, U.S. Bureau of Labor Statistics, RBC GAM

There is also now mounting evidence that domestic American firms are absorbing a fraction of tariff costs as products pass through their hands:

• Q2 earnings calls featured more discussion of margins, which were down from year ago levels in all industries aside from tech and financials.

• Q2 GDP data showed a decline in per-unit non-financial corporate profits driven by higher production and import taxes (likely tariffs).

• August’s producer price data indicated a sizeable 1.7% decline in wholesale and retail margins, led by machinery and vehicle wholesalers.

• Several regional Fed surveys suggest most businesses don’t plan to fully pass tariff costs onto customers.

• The latest National Association for Business Economics (NABE) business conditions survey shows profit margins of goods producers are shrinking as widespread increases in material costs outpace price increases (see chart). Respondents expect this dynamic to continue over the next three months.

U.S. goods producers report margin compression amid rising input costs

As of Q2 2025. Sources: National Association for Business Economics, Macrobond, RBC GAM

Some of this margin compression could be temporary. Businesses might adjust their pricing strategies as they get greater clarity about the persistence of tariffs, perhaps once the Supreme Court rules on the legality of IEEPA tariffs. A Richmond Fed survey found one-quarter of manufacturing firms haven’t raised prices due to tariffs but intend to. Another 47% have already raised prices and plan to do so again.

If domestic profit margins remain under pressure, there is a risk that businesses will resort to layoffs to rein in costs. The NABE survey points in that direction with weakening employment intentions alongside lower margin expectations (see chart below). Indeed, goods producers have shed jobs in each of the past four months, while wholesale and retail employment have also been flat to lower over that period.

We might also expect margin headwinds to restrain business capital spending, but survey responses so far suggest otherwise. But there are multiple competing forces: the jump in CapEx intentions in Q2 likely reflects accelerated depreciation incentives in the recently  passed One Big Beautiful Bill. More broadly, AI-driven CapEx spending continues to act as a powerful tailwind for the U.S. economy.

U.S. goods producers expect margins and employment to decline, but not CapEx

As of Q2 2025. Sources: National Association for Business Economics, Macrobond, RBC GAM

The bottom line is that margin compression might spare consumers from the full brunt of higher prices, but it isn’t great for investors. It might also boomerang back to consumers via layoffs and diminished wealth effects.

-JN

Economic miscellany

Economic resilience

For all of the government shutdown damage and tariff drags, the reality is that the U.S. economy has continued to move forward with surprising vigour.

The Atlanta Fed’s GDPNow model is tracking quite remarkable annualized real GDP growth of +3.8% for the third quarter of the year, and that comes after an identically impressive +3.8% increase in the second quarter. Yes, the first quarter of the year was considerably weaker, somewhat flattering these numbers. But Q2 and Q3 are not just mildly exceeding expectations. They are outright strong, running at around triple the rate that most forecasters had assumed when tariffs were announced.

Productivity pickup

The exception to this rosy economic story has been lackluster job creation (at least when it was possible to get the numbers pre-shutdown). We have no desire to celebrate the lack of new jobs, even if it is less egregious than it first looks after one adjusts for the sharp deceleration in population growth due to lower immigration.

But there is a silver lining when GDP growth is strong and job growth is low. This means that productivity growth – the only other thing that mathematically contributes to an expanding economy – must have been quite good. Indeed, Q2 productivity rose at a +2.6% annualized pace – twice as fast as the average pace sustained since 2010. Higher productivity is the surest means of increasing society’s financial standard of living.

Interestingly, while the second quarter was particularly exceptional, there have been hints of a more enduring productivity pickup dating back to the start of 2023. Annual productivity growth has averaged 1.6% since then, which is pretty good (see next chart). It may or may not be a coincidence that the first major AI large language model was released by ChatGPT at the end of 2022.

U.S. productivity growth remains strong

As of Q2 2025. Sources: U.S. Bureau of Labor Statistics (BLS), Macrobond, RBC GAM

Fascinating job market

In addition to the aforementioned deceleration in economy-wide hiring and a pip of higher unemployment among young workers in AI-affected professions, a number of other interesting trends are afoot in the U.S. labour market.

One of these is that the number of job openings is again below the number of unemployed people (see next chart). To be sure, the unusual thing was that the reverse had been true over the prior several years. The new situation reflects a weakening market, but equally represents a return to normality.

Gap between job openings and unemployment in the U.S. has closed

Unemployment as of August 2025, job openings as of July 2025. Sources: BLS, Macrobond, RBC GAM

In trying to divine the meaning of the still-elevated level of job openings versus the historical norm, it may be that job openings are simply structurally higher than in the past. In this internet age, the cost of posting a job opening is low. Thus, companies may be posting job openings on the off chance that the perfect candidate applies, with no serious intention of hiring anyone if that special person doesn’t walk through the proverbial door.

Another curious aspect of the U.S. job market is that the rate of hiring and firing are both low (see next chart). With the acknowledgement that most of the observations in this chart sit in the murky middle, it is theoretically more normal for hiring to be high and firing to be low (bottom-right corner, when the economy is strong), or for hiring to be low and firing to be high (top-left corner, when the economy is weak). To be in the bottom-left portion of the chart, as is presently the case, is unusual.

Low hiring and firing rates limit U.S. job market churn

As at 09/29/2025. Sources: BLS, RBC GAM

We interpret this to mean that companies expect the economy to be weak and so have slowed their hiring, but the economy hasn’t actually been as bad as they had feared, meaning they haven’t been forced to dispose of much labour. It may also be that companies are hording labour after the trauma of struggling to find new workers during the overheating labour market of a few years ago.

-EL

AI boom cushions U.S. economy

At a time when tariffs have begun to drag on U.S. growth, the economy has been substantially cushioned by the artificial intelligence (AI) boom. To be sure, the eventual hope is that AI technologies radically increase productivity, creating a great deal of financial prosperity. There are little hints that this is starting to happen, but it is only tentative.

Instead, the initial economic boost from AI is coming from the massive capital expenditures many large technological firms are making to build out their models – new chips, new data centres, and so on. Not to be underestimated, though not the subject of this note, enthusiasm about an AI-centric future has also propelled AI-connected equities higher, creating a positive wealth effect for investors that may also be supporting consumer spending.

The so-called Magnificent 7 American technology firms are at the heart of the AI race. The Magnificent 7 are now generating a large 28.7% of all the capital expenditures happening in the S&P 500. They have delivered a stunning 96.5% of the growth in S&P 500-associated CapEx since last year (see next chart).

Magnificent 7 generates large share of CapEx in S&P 500 Index

As of 10/07/2025. Sources: Bloomberg, RBC GAM

Magnificent 7 capital expenditures are on track to rise 58% in 2025 relative to 2024, totaling a gargantuan US$365 billion or 1.3% of overall economic output (see next chart). Many other companies are investing heavily in AI as well, meaning the true AI investment share is likely even higher. That’s not up to the scale of residential investment (4% of GDP), but it’s the same order of magnitude. This is now a segment of the economy that is large enough to move the economic needle.

Magnificent 7 CapEx reflects heavy investment in AI

As of 10/08/2025. Sources: Bloomberg, RBC GAM

These large sums are very much showing up in the relevant economic statistics (see next chart). Accordingly, we posit that AI investment is on track to add at least half a percentage point to U.S. 2025 real GDP growth. That’s a large number in the context of an economy that might normally expect to grow by about 2% per year. It’s even more important in a year when tariffs are theoretically set to subtract a percentage point or more from growth.

U.S. capital investment in computer equipment soared after ChatGPT launch

As of Q2 2025. Shaded area represents recession. Sources: U.S. Bureau of Economic Analysis (BEA), Macrobond, RBC GAM

Looking to 2026 and beyond, AI investment is set to become even larger. Equity analysts project a further 21.5% increase in Magnificent 7 CapEx in 2026 and then 8% growth in 2027. While these will constitute record outlays, the rate of growth is nevertheless substantially slower than that achieved in 2025. In turn, while AI investments should still be contributing to growth, this contribution is set to be smaller than in prior years.

To be sure, estimates vary: Nvidia’s CEO is much more optimistic, anticipating further exponential growth ahead. Perhaps there is some upside risk to the base-case outlook. Conversely, there could be a massive amount of misallocation of capital going on right now that erodes the economic impact of that spending. No one quite knows.

But it remains most likely that the pace of AI-associated CapEx growth starts to slow. In turn, we cannot count on as large a boost from AI for 2026 GDP growth as in 2025. Fortunately, we figure the pain from tariffs will be starting to fade by then, and Fed rate cuts should allow for more economic growth via other mechanisms.

Finally, let us not lose sight of the fact that the main purpose of AI is to boost efficiency. That’s distinct from the mechanical additional of capital spending to GDP. We budget for faster productivity growth in the decades ahead, in large part due to AI. The challenge is in knowing whether annual productivity growth might realistically accelerate by a relatively modest 0.2 percentage points, versus an era-defining +2 percentage points, versus – in a science fiction-like scenario – +20 percentage points per year.

Which countries have room to grow?

In a weird way, long-term economic forecasting is frequently easier than short-term forecasting. Instead of being whipsawed by policy surprises and idiosyncratic economic shocks, the focus is instead on more fundamental and generally stable forces such as population growth and productivity growth.

For investors, long-term economic forecasts are a useful input for strategic asset allocation decisions. They hint at where prosperity may rise most quickly, and in turn where corporate profits might have greater upside and the risk of defaulting on public debt is lower.

Current prosperity is a key theoretical determinant of which countries are poised for the fastest sustained economic growth. For countries with low present income levels, the idea is that they have more room to catch up to living standards elsewhere – simply by importing best practices and adopting existing technologies, and perhaps by lifting educational standards toward global norms. In contrast, countries that already enjoy high levels of income have a theoretically more difficult path ahead: having to invent new technologies from scratch and grappling with potentially diminishing returns on rising educational achievement.

How well does this theory work in practice (see next chart)? This scatterplot of the world’s 50 largest economies confirms a negative linkage between projected economic growth over the next five years and current income (real GDP per capita). Put more simply, on average, poorer countries can expect to enjoy faster economic growth than richer ones as they benefit from the convergence effect.

Poorer countries may enjoy faster real GDP growth

As at 08/25/2025. Includes largest 50 countries in the world by Real GDP; IMF growth based on 2025-2029 forecast. Sources: U.S. Central Intelligence Agency, IMF World Economic Outlook, RBC GAM

However, it must be emphasized that the linkage is not overpoweringly strong. As is apparent in the chart, not all poor countries grow quickly, and not all rich countries grow slowly. Factors such as corruption and the quality of government institutions also have a lot to say about the outcomes.

In response to that importance nuance, the following chart clusters countries not just by their income levels, but also by their real-world prospects. Thus, we can distinguish between the poor nations with strong growth outlooks and those with poor ones.

Linkage between country income and growth is not particularly strong

As at 08/25/2025. Includes largest 50 countries in the world by Real GDP 2024. IMF growth based on 2025-2029 forecast. Sources: U.S. Central Intelligence Agency, IMF World Economic Outlook, RBC GAM

At the most appealing end of the investment spectrum is the low income / high growth group – those countries that are, if anything, out-delivering on the idea that poor countries have a lot of room to grow quickly. These include India, Bangladesh, Philippines, China, Vietnam and Indonesia. Ukraine also appears, presumably reflecting optimistic projections from the International Monetary Fund (IMF) of reconstruction after the war with Russia concludes.

Next are a handful of mid-to-high income countries with mid-to-high economic growth prospects – an attractive combination. This group includes the United Arab Emirates, Romania, Poland, Malaysia, Turkey and Kazakhstan.

Following that group are the emerging market countries in the low-income / mid-growth category. These broadly align with theory, with growth expectations somewhat faster than for their developed world compatriots, but not to an extraordinary degree. The list includes Argentina, Egypt, Colombia, Thailand, Pakistan, Brazil, Algeria, Chile and Peru.

Next is the developed world, with high income levels and relatively slow growth. That said, there are countries within this group that materially beat the average such as Taiwan, Czechia, South Korea and Hong Kong.

Finally, there is the beleaguered low income / low growth group. This is the group that is manifestly failing to deliver on its promise, with no expected economic catch-up in the years ahead. These include South Africa, Iraq, Iran, Nigeria and Mexico.

-EL & SK

International themes

Slow progress on Draghi reforms

It has now been a year since former Italian Prime Minister and European Central Bank President Mario Draghi’s landmark 400-page report on European competitiveness.

The Draghi report called on policymakers to address the European Union’s (EU’s) persistent productivity shortfall by boosting innovation, competitively decarbonizing the economy, and diversifying and strengthening supply chains. It also recommended cutting red tape, addressing skills gaps, and improving policy coordination and removing barriers within the EU’s Single Market. The IMF estimates intra-EU trade barriers are equivalent to an astonishing 44% tariff on goods and 110% on services.

One year on, the European Commission touts progress in several priority areas, including major investments in AI, defence, and cleantech manufacturing and industrial decarbonization. But Draghi himself says progress has been too slow – a view that is backed up by analysis from the European Policy Innovation Council. It found only 11% of Draghi’s reforms have been fully implemented (see next chart).

Few of Draghi’s reforms have been fully or partially implemented

As at 09/04/2025. Sources: European Policy Innovation Council (EPIC), RBC GAM

The think tank said no sector has seen the majority of its recommendations put into action. Transportation and critical raw materials have made the most progress thanks to efforts to secure supply chains and accelerate the electric vehicles (EV) transition. Energy and digitalization/advanced technologies are lagging amid political sensitivity and regulatory complexity.

In a separate analysis, Deutsche Bank noted greater progress in areas where external pressure is most significant, like defence spending, or where national interests are aligned, like cutting red tape. Changes to public procurement and merger guidelines, which are expected to be fleshed out over the next year, could advance the reform agenda. But the challenging political backdrop in some countries – see below – is hampering structural reforms that need to be undertaken at the national level.

Overall, while we are budgeting for some boost to European growth from rising defence spending and German infrastructure investment, it still seems too early to lift our assumptions about the bloc’s long-term productivity growth. But if the large and amorphous “In Progress” category in the chart above starts to materially shift toward the “Implemented” or “Partially Implemented” buckets, that would alter the calculus.

Political instability persists in France and Japan

France just lost its fourth Prime Minister in a little over a year. Sebastien Lecornu resigned after less than a month in office when a key coalition partner objected to some cabinet picks. But French President Macron has not gone far in his search for the next PM, opting to reappoint Lecornu. To remain viable, the government will likely have to court left-leaning parties to a greater degree.

Building a coalition with left-wing parties could come at the cost of pausing Macron’s signature pension reforms that were passed in 2023 to help rein in budget deficits. Recall, Lecornu’s predecessor François Bayrou resigned after lawmakers rejected his budget that called for fiscal consolidation. Appetite for fiscal restraint is limited and the path of least resistance appears to be larger budget deficits.

Political gridlock and deteriorating public finances have caused 10-year French government bonds to trade at slightly higher yields than their Italian equivalents. That’s a sharp reversal of the wider Italian spreads that characterized the post-Global Financial Crisis and euro crisis era. But this is perhaps warranted with France in a slightly worse position than Italy in our fiscal health scorecard.

Meanwhile, Japan’s Liberal Democratic Party (LDP) – the largest partner in the current ruling coalition – elected a new leader and presumptive Prime Minister in Sanae Takaichi. As a disciple of former PM Shinzo Abe and advocate for fiscal stimulus and dovish monetary policy, her unexpected win caused Japan’s yield curve to steepen with a sharp selloff in long-term bonds. Japan already laid claim to the greatest increase in 30-year yields in the G7 year-to-date.

Long-term government bond yields have gone up globally

As of 10/09/2025. Sources: Macrobond, RBC GAM

However, Takaichi’s appointment as PM is now in jeopardy after the LDP’s junior coalition partner backed out of its longstanding alliance. If Takaichi can’t repair that relationship, she could seek other coalition partners, potentially at the cost of even more fiscal handouts. Alternatively, a coalition of opposition parties could appoint its own PM, but that might prove unstable given the LDP’s plurality in the lower house.

Broadly speaking, political instability and fiscal concerns, which are not exclusive to France and Japan, have put upward pressure on government borrowing costs, dampening efforts to provide monetary stimulus via low short-term rates. That said, tight credit spreads and strong equity markets have helped to ease overall financial conditions.

-JN

Skilled worker crackdown

The Trump administration continues to tighten U.S. immigration policy, mostly recently by adopting a more restrictive approach to visas for skilled workers. By increasing the fee for new H-1B visas to $100,000 from $1,500 previously, the administration aims to address some of the program’s perceived shortfalls and limit its use to more experienced, highly skilled foreign professionals.

Next to temporary visas for agricultural workers, the H-1B visa for skilled workers is the most heavily used temporary (nonimmigrant) work permit. The program issued 220,000 visas in 2024 (including renewals). Program intake is limited to 85,000 new visas per year, but demand is much greater with 480,000 applications in FY 2025.

Applicants generally need at least a bachelor’s degree and must work in an occupation requiring “highly specialized knowledge,” though that includes a wide range of fields. More than 40% of H-1B visas go to workers in the professional, scientific and technical services industry. This is a broad category that includes legal, accounting, design, programming, consulting and R&D services among others.

Tech jobs are the most prevalent, and companies like Microsoft, Meta, Apple and Google are among the largest users of the H-1B program (see table below). Other major employers include consultancies (Tata, Cognizant, Deloitte), retailers (Amazon, Walmart) and banks (J.P. Morgan). Remarkably, about two-thirds of H-1B visa holders come from India, while China is also a major source.

As of 09/29/2025. Data for FY 2025 except Source Country (FY 2024). Sources: U.S. Citizenship and Immigration Services, U.S. Bureau of Consular Affairs, RBC GAM

The program was introduced in 1990 to fill skills gaps in the domestic labour force. But critics argue it is being used excessively to the detriment of American workers. Commonly cited issues include:

Undercutting domestic wages: Employers are required to at least match the wages of similarly qualified workers or prevailing wages based on the position and geography. But there is limited government oversight. Researchers at the Economic Policy Institute found that disproportionate fraction of H-1B workers were paid below the median for their occupation and region, arguing that these workers were not especially exceptional and might also be depressing industry wages.

Not filling specific skills gaps: The prevalence of consulting and outsourcing firms among the top H-1B employers suggests the visas aren’t being used to fill specific skills gaps at applicant firms. And with significant excess demand and a lottery system for allocating the visas, they aren’t necessarily being awarded to employees with the most specialized skills or employers with the greatest need.

Leaving fewer opportunities for Americans: The program’s intake of entry-level workers combined with today’s particularly challenging job market for new grads has heightened concerns that H-1B visas reduce opportunities for domestic workers.

The unemployment rate for recent college graduates has historically been below the national average but is now almost 1 ppt higher (see next chart). (As discussed in a recent #MacroMemo, AI and a low-churn labour market are also creating challenging conditions for younger job searchers.)

The jobless rate for all college graduates has also been rising slightly faster than the national average. Unemployment rates are particularly high for graduates with majors in computer engineering (7.5%) and computer science (6.1%) – degrees that align with the tech roles commonly filled by H-1B visas.

U.S. unemployment rate is now higher for recent college grads

As at 08/01/2025. Sources: Federal Reserve Bank of New York, RBC GAM

That said, the program has its supporters. One study found that an increase in the share of H-1B workers within an occupation was associated with a decline in the unemployment rate for that occupation. Another found restrictions on H-1B visas caused multinational firms to outsource jobs – for every visa rejection, they hired 0.4 workers abroad. That tech leaders like Satya Nadella and Elon Musk once received H-1B visas to work in the U.S. anecdotally supports relative openness to skilled foreign workers.

Not unusually for Trump’s controversial policies, the H-1B fee hike is expected to be challenged in court. Visa fees are typically only meant to cover administrative costs, and the existing fee structure was set by Congress, suggesting this could constitute executive branch overreach.

But if the change holds up, it’s worth thinking about the potential impacts. That will depend on the price elasticity of H-1B visas – that is, the extent to which the higher cost reduces demand. H-1B visa holders are allowed to work in the U.S. for an initial 3-year period which is generally extendible by another 3 years. Amortizing the $100,000 upfront fee over 6 years amounts to about $17,000 annually – plus any additional legal fees – if visa holders maximize their stay. That’s less than 20% of the median salary for new H-1B recipients.

If that incremental cost still leaves demand close to or above the 85,000 annual cap, we expect visas to be allocated more efficiently to higher skilled, more experienced workers – which could be somewhat supportive of U.S. productivity. But if demand is curtailed well below the cap, significantly fewer skilled workers may enter the U.S., capping growth – even if those that enter are more experienced on average. A more thoughtful policy might attempt to determine the legitimate need for H-1B visas, set an appropriate cap, and charge a fee that aligns supply and demand.

In either case, less supply of foreign workers could slightly reduce unemployment and support domestic wages, at the margin. But it could also accelerate AI adoption or outsourcing of entry-level jobs to other countries, dampening any positive effects on the domestic labour market.

Would-be H-1B applicants might enter the U.S. using other temporary work permits, like the L-1 (intracompany transfers for managers/executives or employees with specialized knowledge) or O-1 visas (individuals with extraordinary ability or achievement). Both programs are uncapped – 72,000 of the former and nearly 20,000 of the latter were issued in 2024.

As with other aspects of the U.S.’s growing isolationism, other countries stand to benefit from the Trump administration’s more restrictive stance toward skilled workers. To the extent that employers simply outsource jobs to potential applicants’ home countries, India might benefit disproportionately. Others will aim to attract skilled foreign workers shut out of the U.S. – for instance, China recently launched a K visa that allows foreign STEM graduates to apply for residence without a job offer.

Canada previously had success drawing H-1B workers from the U.S. – its Tech Talent Strategy let 10,000 existing visa holders work in Canada for 3 years, but the program was only in place for a year. Following the H-1B fee hike, Canadian Prime Minister Carney indicated the government is evaluating ways to attract tech talent that would otherwise have been eligible for a U.S. visa.

-JN

Canadian recession concerns

The Canadian economy is struggling under the weight of new tariffs and concerns about what the future might bring.

The labour market has been especially soft, with an unemployment rate that has increased by half a percentage point since the start of the year and now rests about a percentage point higher than the level consistent with full employment. Hiring has been choppy in recent months, with big monthly losses interspersed with big monthly gains.

But, on a trend basis, there have been more job cuts than hiring (see next chart). Even accounting for Canada’s nearly non-existent population growth right now, this is poor.

Canadian hiring choppy, but trend is weak

As of September 2025. Sources: Statistics Canada, Haver Analytics, Macrobond, RBC GAM

This Canadian labour market weakness has been concentrated in areas of the economy most reliant on U.S. demand (see next chart). While the decline has been particularly acute in 2025, when tariffs have been repeatedly threatened and implemented, it should be conceded that these industries had been modestly underperforming over the prior few years as well.

Employment has dropped sharply in sectors with significant reliance on exports to U.S.

As of July 2025. Sources: Statistics Canada, Macrobond, RBC GAM

With regard to economic output, the economy shrank outright in the second quarter of this year, raising concerns that the country might stitch two consecutive quarters of decline together. This is a popular rule of thumb for a recession.

In practice, defining a recession is somewhat more nuanced than that. Material declines in output and employment are required, and sufficient breadth of weakness across sectors as well.

Fortunately, while the circumstances are clearly challenging for Canada, there are some silver linings.

1. Second-quarter Canadian GDP had weak foreign demand, as expected. But domestic demand actually rose quite nicely (see next chart). Within this, consumer spending increased with surprising enthusiasm, especially when measured on a per-capita basis (see subsequent chart).

Canadian real domestic demand rose in the second quarter of 2025

As of Q2 2025. Sources: Statistics Canada, Macrobond, RBC GAM

Canadian real consumer spending per capita shows rising enthusiasm

As of Q2 2025. Sources: Statistics Canada, Macrobond, RBC GAM

1. Third-quarter Canadian GDP is presently tracking a modestly positive number after a good-looking GDP increase in July (+0.2%) and a flash estimate that August was approximately flat (see next chart). So, two consecutive quarters of decline are unlikely.

Latest Canadian GDP tracks a modestly positive number

As of August 2025. Sources: Statistics Canada, Macrobond, RBC GAM

The tariffs on Canada aren’t coming in as badly as feared, with a current effective rate of just 7%. That’s among the lowest tariff rates in the world, though of course still painful given such deep trade ties with the U.S.

The Bank of Canada is still likely in a position to cut rates again before the end of the year – whether October 29 or December 10 – but this helps to explain why the rate cutting has not been to the bone.

-EL

USMCA expectations

The nearly existential question for Canada and Mexico, and a question of some importance for the U.S., is what happens to the trilateral USMCA (U.S.-Mexico-Canada Agreement) deal when it comes up for renewal in mid-2026. The following discussion is framed from a Canadian standpoint.

Even with the existing USMCA deal nominally in place, Canada is being hit by significant U.S. sectoral tariffs. As it happens, Canada is the largest exporter of steel, aluminum, copper and lumber to the U.S. (see next four charts). Each of these products has been subjected to hefty Section 232 tariffs.

Canada led exports of steel products to U.S. in 2024

Sources: U.S. Census Bureau, Macrobond, RBC GAM

Canada also led aluminum exports to U.S. in 2024

Sources: U.S. Census Bureau, Macrobond, RBC GAM

Canada led exports of copper/copper articles to U.S. that are subject to 50% tariffs

Copper imports of ores and concentrates, mattes, cathodes, anodes and copper scrap are exempted. Sources: UN COMTRADE, Macrobond, RBC GAM

Canada led in lumber exports to U.S. in 2024

Includes only lumber exports subject to 10% tariffs effective 10/14/2025. Sources: U.S. Census Bureau, Macrobond, RBC GAM

Conversely, if the Supreme Court were to overturn IEEPA (country-level) tariffs, Canada and Mexico would be relatively less affected since only a small fraction of the two countries’ traded goods are presently subject to those tariffs.

Expectations are rising that there will be separate U.S.-Canada and U.S.-Mexico bilateral deals alongside the trilateral USMCA, so as to address different U.S. concerns for each country. This also constitutes a divide-and-conquer strategy for the U.S., as it has better leverage in one-on-one negotiations than one-on-two.

There are also reports that the Trump administration wants to specifically punish Canada for its earlier tariff retaliation. There is considerable irony to this as Mexico is by far the greater offender when it comes to fundamental U.S. complaints such as the cross-border flow of illegal drugs and undocumented immigrants, the size of the country’s trade surplus versus the U.S., the undercutting of U.S. labour costs, its rising share of North American auto production, and the transshipment of products from China.

The U.S. continues to seek a range of concessions from Canada as negotiations proceed (see next table). Canada has already taken significant action, including removing its digital services tax and eliminating the global minimum tax for U.S. firms. Canada is also in the process of increasing border security and has committed to substantially increase military spending.

Potential U.S. demands of Canada

As at 10/14/2025. Source: RBC GAM

The question is how much further the U.S. will push. As the table illustrates, the U.S. has floated many other demands. There appears to be particular focus on joint U.S.-Canada military initiatives including the proposed Golden Dome missile defence system for North America, getting Canada to buy more U.S. goods and to invest more into the U.S., and gaining greater access to Canada’s supply management industries.

Canada likely does join the Golden Dome, and very likely does commit to additional U.S. purchases/investment. Prime Minister Carney recently spoke of Canada investing US$1 trillion into the U.S. over the next five years. How the accounting for this works, who will do the spending and whether it is truly new money or simply funds that would have flowed anyway is open for debate. Many other countries have made similar commitments with similarly unclear accounting and implementation plans.

It seems reasonable to expect the existing sector-oriented tariffs to persist and to be formalized within the USMCA or an accompanying bilateral deal. This may be tempered by product-specific export quotas allowing a lower tariff below that threshold.

Auto assembly is a particularly large risk, as Commerce Secretary Lutnick has said that the U.S. is no longer interested in buying Canadian-made cars. That suggests there could be a greater barrier to auto trade erected than the current arrangement of tariffing the non-U.S. value-added share of imports. The auto parts sector looks to be under somewhat less direct threat, though assembly and parts are of course closely linked.

A major unknown – and danger – is whether the USMCA veers into service sectors. It is difficult to tariff services, but the U.S. is already seeking to tariff foreign-made films, which would represent a major step in that direction. Canada’s financial sector has been the target of occasional White House complaint, and Mexico’s telecom sector has also received negative attention.

One wonders whether the U.S. will attempt to secure easier access for American firms to key service industries in those countries. However, it should be noted that the equivalent U.S. sectors are similarly regulated/protected.

On the net, Canada and Mexico are likely to emerge in a moderately worse position than they were going into the USMCA negotiations. They will probably have to make additional (mostly) non-trade concessions and accede to the formalization of tariffs on key industries into a legislated trade document – as opposed to the existing executive orders that might easily be reversed by a future U.S. administration.

Conversely, quotas might allow some Canadian and Mexican goods in targeted sectors to flow into the U.S. at a somewhat lower tariff rate. We do not budget for more extreme action, such as in the services direction, though it is possible.

-EL