Month: June 2023

I might as well get this over with. Until a couple of weeks ago I was responding beautifully to the CAR-T treatment, with my cancer load dropping smoothly toward zero. Then it all stopped, with my M-protein level stalled at 0.34:

This is a "partial response," and for all practical purposes it means the treatment didn't work. There's a good chance it will keep me chemo-free for a few months or a year, but that's about it.

This hit me out of the blue. I was convinced everything was going well and it barely even occurred to me that my latest result would be anything but zero. Needless to say, this is pretty depressing news.

According to the Society of Family Planning, the abortion rate nationwide has dropped 3.3% since last year's Dobbs decision. Here are the states where the abortion rate went up and down the most:

Texas went from nearly 3,000 abortions per month to zero. Florida, where abortion bans remain mired in court, saw an increase because it's close to lots of Southern states that enacted strict abortion laws. States like California and New York saw fairly large increases in the number of abortions they provided, but their numbers weren't all that high on a percentage basis.

UPDATE: Note that these numbers do include most medication abortions using pills like mifepristone—but only if clinical assistance is involved (either in-person or online). Medication abortions aren't counted if they're done entirely outside the health care system. It's impossible to say if this affects the final figures either up or down.

Jeez, another reasonable decision:

States can’t use the federal courts to try to force the federal government to arrest and deport more people who are in the country illegally, the Supreme Court ruled Friday.

The 8-1 decision could cut down on a flood of lawsuits recent administrations have faced from state attorneys general and governors who disagree with Washington on immigration and crime policy.

This is a decision based on standing, but it sure seems to address the merits of the case too:

The States essentially want the Federal Judiciary to order the Executive Branch to alter its arrest policy so as to make more arrests....[But] the Executive Branch—not the Judiciary—makes arrests and prosecutes offenses on behalf of the United States....That principle of enforcement discretion over arrests and prosecutions extends to the immigration context.

....The Executive Branch must prioritize its enforcement efforts. That is because the Executive Branch (i) invariably lacks the resources to arrest and prosecute every violator of every law and (ii) must constantly react and adjust to the ever-shifting public-safety and public-welfare needs of the American people....That reality is not an anomaly—it is a constant.

....If the Court green-lighted this suit, we could anticipate complaints in future years about alleged Executive Branch under-enforcement of any similarly worded laws—whether they be drug laws, gun laws, obstruction of justice laws, or the like. We decline to start the Federal Judiciary down that uncharted path.

That's clear enough. The decision was satisfyingly 8-1, with (of course) a red-faced Sam Alito offering the only dissent. "It renders States already laboring under the effects of massive illegal immigration even more helpless," he howls—before pivoting out of nowhere into a disquisition on "regal authority" and the overthrow of James II. That's our Sam.

The "smart Trump" speaks:

Question: Are you in favor of eliminating any agencies?

DeSantis: We would do education, commerce, energy, and the IRS. If Congress won’t go that far, I’m going to use those agencies to push back against woke ideology… pic.twitter.com/cAoZIXfESu

— Acyn (@Acyn) June 28, 2023

Eliminating cabinet departments is the last refuge of idiots. DeSantis isn't going to "do" any of them, nor does it matter unless he's also planning to eliminate all the actual programs they run. I wonder if he even knows that the Energy Department runs R&D for all our nuclear weapons and the Navy's nuclear reactors? Or that someone has to collect taxes whether we like them or not?

Moron. Either put names to the actual programs you plan to kill or STFU.

This is the ornate, Beaux-Arts style Alexander III Bridge crossing the Seine in Paris. I was strolling around early when I took this and caught it just as the morning sun was shining directly on the winged horses at the top of the pedestals. The Grand Palais is in the background.

When you look at some sort of domestic trend or other, it's often helpful to look also at the rest of the world. This was one of the things that tipped me off about the connection between lead and crime, for example. Everyone knew that crime in the US had dropped during the '90s, and this had prompted a rash of theories to explain it: drugs, policing, abortion, broken windows, a growing economy, etc. But it turned out that crime was also down in Canada. And Germany. And Britain. And Australia. And all over the rest of the world. Clearly something had to be going on that wasn't unique to the US.

Today, the CEA does a similar exercise for inflation in the G7 large economies (mostly Europe). Here it is:

This got me curious about smaller, less central economies. So I grabbed a few inflation rates from other continents:

Obviously there are differences, but not big ones. What this means is that inflation in the US almost certainly has little to do with domestic policies like the stimulus bill or easy money from the Fed. Inflation is a global phenomena, hitting everyone at roughly the same time and then declining at roughly the same time.

In other words, it's COVID. It's supply chains. It's shortages. It's oil prices.

There can still be idiosyncratic explanations for parts of the picture. In the US, eviction moratoriums at the start of the pandemic pushed up rental prices. Stimulus spending was higher than in many places and probably added a point or two to the inflation rate temporarily.

But these are small things. In a nutshell, it's COVID. It's supply chains. It's shortages. It's oil prices. Those things are all going away naturally, and inflation will shortly go away along with them.

I've long argued that it takes about a year for Fed interest rate hikes to slow down the economy and affect inflation. This means that Fed hikes have had little impact so far, but will start to have one over the coming year. Jason Furman reviews the evidence and he isn't buying it:

In sum, just about every way monetary policy affects the economy reached its current level of tightness over a year ago. Just about everything you would expect to be affected by monetary policy (housing, nonresidential structures, net exports) was--but in the past.

— Jason Furman (@jasonfurman) June 28, 2023

Furman's argument is fairly standard: Fed interest rates don't directly affect the economy. They affect it mostly by pushing up long-term rates, which in turn affect housing, investment, commercial loans, and so forth. That initially happened via expectations even before the Fed started hiking and had finished its work by the end of 2022. So tightening is all in the past. There's something to this:

The Fed started raising interest rates in March of last year, but longer-term rates had already started to rise a little bit a few months earlier. These rates continued to climb through October but have since mostly flattened out. Their job is done.

Can this be right? If it is, it means that "long and variable lags" these days take about six months from the very first hint of interest rate hikes. Hell, the Fed has continued raising rates by a couple of points since October, but if Furman is right this has had no effect at all. Markets had long since incorporated that expectation and simply didn't respond.

So could the Fed have just gone on vacation starting in October? Was the increase since then from 3% to 5% necessary? Would any further hikes have any impact?

The best I can say is that although this affects my beliefs, I don't think it overturns them entirely. Empirical evidence still suggests strongly that Fed hikes take longer than this to work their way fully through the economy, especially when it comes to inflation—which is one of the most stubborn economic measures to respond to tightening.

So if I had to update my priors, I'd now say that (a) Fed hikes probably started affecting inflation a few months ago, earlier than I would have thought, but (b) there's still tightening in the system that will reduce inflation for the rest of the year. We don't need any more.

[WARNING: This post is almost certainly wrong. Click here for an update showing how and why Mississippi test scores really did go up after 2013.]

I think I may have been wrong about something.

Over the weekend I wrote a couple of posts about Mississippi's "reading miracle." I looked things over and decided the evidence pointed toward genuine improvement following a 2013 reform that emphasized phonics instruction. Before the reforms Mississippi's white kids scored nine points below the national average on the NAEP test. By 2022 they were scoring five points higher. Black kids did about the same.

Not bad! But one of Bob Somerby's comments continued to niggle at me. Among other things, the 2013 law mandated that third-grade kids who failed a year-end reading test be held back a year. This affected about 9% of all third graders.

This means that if we then test in fourth grade we're automatically going to get higher scores than we should because the bottom kids are no longer in the testing pool. They're still back in third grade. I figured this effect would be small, but I figured wrong:

In 2013, Mississippi fourth graders are 13 points below the national average if you look at all students. After that year, "all students" excludes all the bottom kids who were held back. What we need to know is what the Mississippi scores would look like if we added them back in.

By chance, this turns out to be fairly easy to estimate, so I did that. It's the orange line. Follow it along and it turns out that Mississippi scores in 2022 are still about 13 points below the national average. In other words, the 2013 reforms had all but no effect.

Easy come, easy go.

Over at the American Prospect, David Dayen argues that Democrats need to deliver the goods if we want people to vote for us. That is, we need to adopt a deliberate and sustained policy of "deliverism." The problem is that we think too small. Take Obamacare, for example:

First, the goal of Obamacare was to insure more people, and it did. Roughly 85 percent of Americans had health insurance in 2008. Today, it’s about 90 percent.

....In 2009, the average medical cost for a family of four was $15,609. Today, it’s $30,260. That’s almost the cost of a new car in health care costs, every single year. In other words, 85 percent of potential voters have the same or a worse experience with health care today, versus 5 percent who gained insurance. It’s hard to call that a net economic improvement in the lives of most voters.

This isn't quite fair. David is right about the uninsurance rate but off by quite a bit when it comes to how much people spend on health insurance:

The uninsured rate has dropped about six points under Obamacare, which represents something like 15 million people. But when you analyze overall health care spending, what matters isn't the total cost of health care but how much you have to personally shell out from your own paycheck. That's out-of-pocket deductibles plus your share of premiums, and it comes to about $3,000 these days. As you can see, Obamacare really did help to push this down. Most people are having a better experience with health care.

Nonetheless, there's no denying it: six points in the uninsurance rate isn't a lot. And while the growth of health care costs has slowed for the average person, actual costs are still going up. Hardly anyone is going to notice that, and they certainly aren't going to base their vote on it.

The other problem, as David says, is that even if a program is big enough to be noticeable it isn't much good if it's temporary. Here's a fascinating bit of polling about the Child Tax Credit included in the American Rescue Act, which was quite large and noticeable:

I'd be cautious about inferring causality from a very few data points like this, but what it suggests is that people who received the CTC liked it and gave Democrats credit for it. In late 2021, 49% planned to vote for a Democrat for Congress. A month later the CTC expired and the Democratic vote went down to 44%. A couple of months later it was a point further down. It's hard to say if Democrats got any credit for this. It's quite possible that they lost more support for "taking it away" than they got in the first place for starting it up.

So what to do? The problem, I think, is that there aren't any programs left that are big enough to really get people's attention. Big early economic programs—Social Security, unemployment insurance, Medicare, Medicaid—as well as big early social programs—the Civil Rights Act, the Voting Rights Act, Americans with Disabilities—got people's attention because they started from nothing and then kept growing in importance year after year. Aside from genuine national health care, there's really nothing left like that. Here are David's suggestions:

Democrats could reverse their toxic image in many parts of the country by reversing policy choices on subjects like NAFTA, deregulation, and banking consolidation, which have helped hollow out the middle class for decades.¹

Well. Maybe. But those don't frankly seem like things that most people even care much about, let alone things that would move the voting needle much. As for "hollowing out" the middle class, it just ain't so:

Over the past half century, middle-class incomes have grown 40%. That's more than working-class incomes and less than upper-middle incomes. To the extent that people have moved around, it's mostly working-class folks moving up into the middle class and middle-class folks moving up into the upper middle.

This is the problem right now—if you want to call it a problem at all. The typical middle-class household earns about $70,000 per year, and it's hard to get the torches and pitchforks going among a populace so comfortable.

To an extent, the problem here is less one of "deliverism" and more one of catastrophism. We Democrats insist on an endless message of economic doom and gloom even though all the evidence in the world suggests that most Americans are fairly well off, fairly satisfied with life, and certainly not ready to revolt. Sure, deliverism works only if it's big and noticeable, but it also works only if voters feel a desperate need for it. Right now very few do. This is why social issues dominate the political landscape and probably will for quite a while.

¹As an aside, I'd add long-term nursing care to this list. It's pretty sizeable and it's a very noticeable problem for a lot of people. Aside from national health care, I'd guess that it's about second on the list of things Democrats could deliver that would really make an impression—assuming of course, that we're able to pass something simple, blunt, and comprehensive. That's what it takes.

Old and new. The top photo is a picture of a Los Angeles metro tunnel on the Red Line—completely modern and bored smoothly into a circle barely larger than the trains themselves.

The bottom photo is a Paris metro tunnel on Line 12. It's old and gigantic, and the stations are so close together you can see one from another. But that's a good thing: you never have to look far in Paris to find a metro station.