On the menu today: As mentioned yesterday, my list of what U.S. government leaders ought to prioritize couldn’t be laid out in one day without the newsletter growing to an epic length. After yesterday’s analysis of our declining birth rate (and its domino effects on various industries, public education, higher education, those entering the workforce, and the consequences of AI), let’s turn our attention to the latter half of life and the problems that most of our governing class would rather not talk about much.
America’s Aging Issues
Yesterday’s edition of this newsletter concluded on the concern that AI is most quickly eliminating entry-level jobs, making it hard for young people to start their careers. But perhaps the AI revolution will come to a premature end, because . . .
6. Our aging power grid will soon start buckling under AI-driven demand.
All that booming investment into artificial intelligence is causing demand for electricity to rise much more quickly than previously projected:
PG&E entered 2026 expecting a year’s worth of new electricity demand. Barely two months later, nearly all of it was already spoken for. Interconnection requests were piling up faster than planners had expected, overwhelming a regulatory system built for an era when electricity demand barely moved.
That world is gone.
Load growth that historically ran below one percent annually hit four percent at some grid operators last year, according to a report by the Lawrence Berkeley National Laboratory. Bain and Company projects that AI data centers alone could consume up to 9% of total U.S. electricity by 2030, adding more than 150 terawatt-hours of demand that the current grid was never really built to handle. A third of that new demand is concentrated in Virginia, Texas, and California, according to Pew Research Center, putting extraordinary pressure on regional systems already straining to keep up.
You can’t exactly snap your fingers and generate more electricity; “typically a nuclear power plant will take over five years to construct whereas natural gas-fired plants are frequently built in about two years.” And that’s just the time to build it once all the permitting issues, environmental reviews, and lawsuits are resolved, which as you’ve probably noticed, can take a long time to sort out.
Here the Trump administration has the right goal in mind, and has taken a couple of steps here and there — mostly executive orders and the declaration of a national emergency, the announcement of pilot projects for nuclear reactors, etc. But the administration also spent $4 billion to cancel offshore wind projects and other renewable energy projects. Some of those canceled projects will result in the production companies shifting funds to liquefied natural gas projects. But it still amounts to canceling projects that have already started in order to start new ones. Once you throw in the time to sort out permitting issue, it usually takes more than four years to build a power plant, and with the change in presidents, the U.S. tends to dramatically alter its energy policies and priorities every four years.
To be useful, a power plant must get connected to the grid, and a July report from Lawrence Berkeley National Laboratory laid out that the permitting and approval process to get hooked up to the grid has a frustratingly long backlog:
Most interconnection requests are never actually built – proposed power plants may not come to fruition for a variety of reasons. Looking historically at interconnection requests submitted from 2000-2020, just 13 percent of capacity had come online by the end of 2025, while 75 percent had withdrawn (final outcome of more recent projects cannot yet be determined). Withdrawal rates are relatively high (>40%) even after signing an interconnection agreement, posing challenges for grid planners and operators. Additionally, interconnection process timelines remain high: the median duration from submitting a request to signing an [interconnection agreement] was well over 3 years in 2025, while the timeline to reach commercial operations exceeds 5 years (for the subset of regions with available data).
Similarly, the ongoing rage about data center construction may well slam the breaks on the expansion of AI; as NR’s editorial notes, arguably the lone U.S. political figure who has resisted the siren call of populist rabble-rousing and demonization of data centers is . . . (checks notes) . . . President Donald J. Trump. Credit where it’s due!
Of course, permits and regulations are only one obstacle; an earlier edition of this newsletter laid out all the physical challenges to building a data center.
7. Social Security’s approaching insolvency.
Perhaps you look at all these worsening issues and shrug that it will be someone else’s problem, because you’ll be retired in the not-too-distant future.
Social Security’s Old-Age and Survivors Insurance Trust Fund will be insolvent in 2032, and the combined Social Security trust fund will run out in 2034. If you’re elected to the U.S. Senate this year, these crises will begin on your watch. As the Committee for a Responsible Federal Budget (CRFB) lays out, “When the Social Security retirement fund runs out of reserves, beneficiaries will face an abrupt 22 percent cut.”
Considering the severity of the consequences and the rapidly approaching deadlines, you would think political leaders would talk about the ticking time bomb of entitlements more. But as the Brookings Institution noticed, there’s arguably less discussion than ever: “There is little sign that would-be senators are paying any attention to the program. Few candidate websites mention the program’s financial difficulties, and only seven out of 56 candidates describe their policy stance on improving Social Security’s finances.” It seems that the worse the problem gets, the more candidates for office don’t want to talk about it.
8. Medicare’s parallel funding cliff.
In 2033, right in between the two impending Social Security insolvencies, Medicare Hospital Insurance is projected to become insolvent as well. CRFB calculates that “Medicare insolvency would lead to an 11 percent cut in payments” to inpatient hospital care, skilled nursing, home health, and hospice care. Doctors already argue that Medicare’s reimbursement rates are far too low; the head of the American Medical Association argues, “It’s a 30-something percent gap between where our costs have gone and where our compensation has gone.”
9. The long-term care workforce shortage.
You basically have two options in life: live long enough to require long-term care or die before then; seven out of ten Americans will require long-term care. Smaller families mean a greater burden on the adult children of seniors, or a need for professional long-term care. Professional long-term care costs an arm and a leg; “The national average cost for a semi-private room in a nursing home is $112,420 per year.” (It would be cheaper to send all our aging seniors back to college.)
Even if you’ve got the money, there’s the question of who’s going to be providing all this long-term care, as, “According to the U.S. Bureau of Labor Statistics, the country will need to fill 9.3 million direct care job openings by 2031.” You might be thinking that at least we’re keeping the long-term care workers that are already here, but in fact we are getting rid of them. As Republican Representative Mike Lawler wrote in June, “Of the 350,000+ lawful Haitian TPS holders, roughly one third work in our healthcare system. Immediately shutting off TPS will create a crisis in our hospitals, nursing homes, and in the I/DD community.”
An April 2025 study found that “more than one million noncitizen immigrants (one-third of them undocumented) work in health care in the U.S. . . . Of 55,802 noncitizen physicians, more than three-quarters worked in hospitals, where they accounted for 8.8 percent of all physicians. A total of 148,445 noncitizens worked as registered nurses (98 536 of them in hospitals) and 443,958 worked as nursing aides or assistants.”
We all want to get rid of the gang members, drug dealers, thieves, child molesters, rapists, and drunk drivers who entered our country illegally or overstayed visas. The national need to get rid of grandma’s nursing home aide is less clear.
10. Overseas looming threats.
Up until now, this list of real problems has focused entirely on domestic issues.
Beyond our borders, there’s the very real threat of a Chinese invasion or blockade of Taiwan, or China getting into a hotter conflict with our Asian allies over territorial claims in the South China Sea.
Earlier this summer, the Stockholm International Peace Research Institute concluded, “China is expanding its nuclear arsenal faster than any other country and showcased several new nuclear systems during its 2025 military parade. By January 2026, China had loaded hundreds of missiles into three large missile silo fields in the north of the country, while working to complete 30 silos in three mountainous areas in the east.”
I don’t know why Xi Jinping wants a much larger nuclear arsenal, but I can’t see any upside in that development for us.
No matter how the Russian invasion of Ukraine concludes, the next leader of Russia after Putin is not likely to be much friendlier to the West. A combination of geography (few natural borders with eastern Europe) and a deep-seated, seemingly intractable paranoia and jealousy of the West mean that for the foreseeable future, the United States will remain the Russian state’s “glavnii protivnik” — main enemy.
I won’t go through all of the various overseas threats — North Korea’s nuclear program, cyberterrorism, the stability of the Pakistani government and who controls that country’s nuclear weapons program, or another shooting war between India and Pakistan.
The U.S. government is hardly ignoring these threats, but as we’ve seen in our efforts in Iran, our ability to influence events overseas is limited. They all seem like higher priorities than threatening to invade Greenland and spurring the Danes to send troops prepared to blow up airfield runways.
(I won’t even mention the use of social media to spread additional panic during terrorist attacks, the engineering of biological weapons to target particular genes, or a breakthrough in quantum computing making all existing computer firewalls obsolete.)
This list is besides the national debt, but I’d like to throw out a bonus eleventh problem that focuses on one aspect of the debt, which is the increasing cost of interest payments on that debt.
11. The national debt interest payment spiral.
As the U.S. General Accountability Office explains, “The government finances annual deficits by issuing federal debt (Treasury bills, notes, and bonds) to domestic and foreign investors (debt held by the public). The government — like businesses and individuals — must pay interest to the investors: this is the cost of borrowing. As debt grows, interest rates can increase, which further increases borrowing costs.”
The more you borrow, the more you have to pay in interest payments in order to not default on that debt.
Last fiscal year, the U.S. paid $970 billion in interest; this year, as of July, we have paid $931 billion, and this will probably be the first year we pay more than $1 trillion in interest. This year, through July, we have spent more on interest than on defense ($806 billion) and almost as much as Medicare ($955 billion). Each year, a larger chunk of the federal revenue coming in — $5.23 trillion last year, $4.48 trillion so far this year, $139 billion more than at this point last year — must be used to make those interest payments.
As the Bipartisan Policy Center observes, a whole lot of government borrowing means higher interest rates for everyone: “When the government borrows more, it absorbs funds that would otherwise finance private investment, and interest rates rise as private borrowers compete for what remains. . . . Rising debt can heighten investors’ concerns about future inflation and even potential default — whether by failure to pay bondholders or by printing money to cover obligations. Lenders respond by demanding higher returns on long-term debt as compensation.”
So, it’s not just that the more the U.S. government borrows, the more expensive it gets for the government to borrow additional amounts; it gets more expensive for everyone to borrow.
On any given day, President Trump can focus his attention on “fake polls,” or the renaming of the airport in Palm Beach after himself, or the White House ballroom, or who Tucker Carlson is meeting with this week, or the scandalous affair of former Representative Eric Swalwell (you know, the guy resigned from Congress in April), or his latest proposal to rename Lake Ontario “Lake America.”
The president is an all-time champion in “majoring in the minor things.”
But he’s not that different from a lot of members of Congress, and a lot of players on the political scene. The problems of fewer babies being born, an insufficient power grid, lack of health-care workers, unsustainable entitlement programs, deterring military invasions — these are big, complicated, difficult, and scary problems that don’t lend themselves well to easy demagoguery. As a result, much of our governing class just isn’t that interested in those topics. And the problem gets a little worse each year.
Tomorrow’s edition of this newsletter will be cheerier . . . by default.
ADDENDUM: Over in the Washington Post, a look at the less-than-enthusiastic response of Republicans to President Trump’s decision to launch another trade war with Canada, just a few months before the midterm elections. Hey, it’s not like Maine, Michigan, Minnesota, New Hampshire, or Ohio rely on trade with Canada much, right? No important Senate races there!