Jan-23 Q&A Part I

Inflation, Automation, & Breakfast Conspiracies

Welcome to Part I of our January Q&A. We got a ton of questions this month, so we split it into two batches. I (Zack) took Part I. Anirban will handle Part II, which should go out in the next week or so. Part II will only be for paying subscribers. If that’s not you and you want it to be, just click the button below:

If you have any last minute questions to submit for Anirban, you can do that here.

Why are food and energy excluded from measures of inflation? Is it because you can’t apply hedonic adjustments to them (i.e., a dozen eggs is just a dozen eggs - there are no new attributes they can tweak to make it seem more expensive when it isn’t).

Inflation measures exclude food and energy because they’re volatile. Just look at the graph below; those prices bounce all over the place. Core inflation (prices minus food and energy) is really steady.

A few points. First, the prominent legislated uses of the Consumer Price Index—like social security adjustments and I-bonds—use overall inflation, not core inflation.

Second, we (the people who write about this kind of thing) like core inflation because it tells us more about future inflation. For instance, if overall inflation is 1.0% for a month (very high) because gas prices spiked but core inflation is just 0.2% (normal, bordering on low), you don’t have a big problem, at least in the long term. If overall inflation is 0.2% because gas prices fell but core inflation is 1.0%, it’s probably time to panic.

How do the shortages/price surges for items like baby formula, bacon, eggs (a breakfast food conspiracy?) impact the larger economic picture?

If there is some nefarious plot to drive up breakfast prices, it’s not having much of an impact on the broader economy. Eggs were just 0.147% of the average consumers expenditures in November 2022 (according to the Consumer Price Index). Bacon was less than 0.1%. Even if these prices skyrocket, it’s not a huge deal.

Unless you’re a baker, or run a breakfast shop, or need baby formula to feed your infant. Then it is a big deal. So on the personal level, these shortages can be really tough, but the rest of the economy is pretty insulated from negative impacts.

(Side note: shortages aside, price increases for eggs and bacon are less of a problem than price increases for baby formula because of the substitution effect.)

Some goods do matter for the broader economy, though. Gas prices, for instance, are more than 4% of an average consumer’s spending, so a rapid increase can have a big impact on finances in the short term. Cars are nearly 9%, and shelter nearly 33%. When those prices go up, it can lead to all kinds of secondary effects.

In school (in the 70s/80s) they said we’d switch to the metric system at some point. Why hasn’t this happened? What would be the economic impact of switching to the metric system

Metrication is a movement that seems to come and go, but mostly went (or maybe I just wasn’t paying attention) during my schooling days in the 1990s and 2000s. Why hasn’t it gotten more traction?

One reason is that we, American adults, don’t want to switch to metric. It would be easy enough to teach it to the next generation. It would be difficult bordering on impossible to get us to learn it.

Another reason is that we have a lot of things already built for imperial. Our speedometers show miles per hour. Our scales and medical records are in pounds. Our thermometers measure in Fahrenheit.

But we’re a lot more metric than you might think. We (the royal we, not me) run 5k and 10k races. Liquor is measured in milliliters, nutrients in grams. Gems are measured in carats, and medicine mostly uses metric, too.

Regarding the economic impact of switching, I think it would be a negative in the short term—every dollar we spend putting up a new speed limit sign is a dollar we can’t spend on other infrastructure—but a positive in the long term as it simplifies international trade, eliminates conversion costs in industries like construction and manufacturing, and reduces mistakes (like the one that led to the loss of the Mars Climate Orbiter).

Switching to metric is almost certainly a good idea. I’m still against it for the sheer inconvenience. Yes, this is lazy, but I bet a lot of Americans feel the same way.

The economic concept at play here, for those interested, is path dependence.

Tech has had layoffs. When do you foresee a meaningful demand decrease to hit the manufacturing or construction industries?

Falling demand has already hit the manufacturing segment. Manufacturing output fell 1.3% in December. Both the ISM and S&P Global manufacturing PMIs show the industry in contraction. Same goes for regional manufacturing surveys from the Fed.

Construction activity is, for the moment, holding up better, but the industry tends to lag the broader economy. During the Great Recession (’08 financial crisis), construction spending bottomed out in February 2011, a full year after the low point for economywide employment. Given high backlog and some pent up demand for construction services from during the pandemic, it’s not clear when/if construction might slump.

Regarding layoffs, I’m really not sure. There’s a pretty severe shortage of workers, and despite lots of high-profile layoffs, the unemployment rate is still at its lowest level since 1969. The demand for workers is so much greater than the supply that it’s conceivable some demand can be destroyed without causing a meaningful increase in layoffs.

That said, I expect next Friday’s jobs report to show the slowest monthly employment growth since we lost jobs in December 2020.

Will illegal immigration help with our tight labor market? Have you included that in your forecast?

This is a tough one because undocumented immigrants are, well, undocumented. Setting aside labor force impacts, we don’t have great data on how the flow of undocumented immigrants is changing over time. Pew Research and DHS put out estimates, but the accuracy is questionable (for what it’s worth, Pew thinks the unauthorized immigrant population peaked in 2007).

As to whether it’s included in our forecast, yes and no. I don’t think many forecasters are specifically modeling the impacts of undocumented immigrants, but unless there’s a really drastic change in the flow of migrants, it’s baked in due to its impact on indicators that are included.

Said another way, we’ve had undocumented immigrants in the workforce and will continue to have undocumented immigrants in the workforce, yet we’re still dealing with labor shortages.

When will the world adjust to the new normal of labor shortages? What does the world of 2033 look like? Robots are already taking over McDonalds... will they also be cutting our hair, changing our oil, and writing our best-selling books? Will automation ever catch up to the declining work force?

It's tough to know what the economy will look like in 2033, especially when we’re busy trying to hash out what the middle of 2023 will look like. One thing I’m really confident about: a robot won’t be cutting my hair anytime soon (there are plenty of nightmarish videos of this).

Automation is, in the long run, a good thing, and I think we’ll continue to automate things wherever possible. We’d still be better off (economically) with faster population growth. Fewer taxpayers supporting more retirees is never a good thing. To the extent we don’t have enough workers to meet the demand for labor, yes, people will increasingly try to automate processes that would otherwise be done by a human.

Some people are really optimistic (or pessimistic, but certain in either case) that recent AI developments will lead to lots of automation. I’m less convinced but could definitely be wrong; a lot of people smarter than I am think otherwise.

One small technical point: when you order at a McDonalds kiosk or use self-scan at a grocery store, it’s not really automation, it’s just the customer doing what an employee used to do. Automation would be a machine that takes your groceries off the belt, scans them, bags them, collects your payment, and gives you a receipt.

This is a really big topic, so here are a few links to more in-depth pieces for anyone interested.

Noah Smith: American workers need lots and lots of robots

Tyler Cowen: AI is improving faster than most humans realize

Greg Rosalsky (talking about superstar economist Daron Acemoglu): A new way to understand automation

What is your opinion on companies starting to require more time in the office. Will employees look elsewhere for jobs with so many openings or head back to the office?

Some people forced to work in the office will definitely look for jobs elsewhere, but that doesn’t mean companies bringing workers back into the office are making a mistake. It seems like the big issue with remote work is onboarding new hires and managing those new hires, and it’s quite possible that some companies (like a Goldman Sachs) have decided they’d rather not have the workers who won’t come in anyway.

I still don’t have a good feel for the steady-state equilibrium on remote work. It’s definitely something to keep an eye on, but I suspect the dust won’t settle for at least another couple years.

Wal Mart was the originator of 24-hour shopping. With the pandemic behind us, why the hell haven’t they re-opened up 24-hour shopping?

My guess? Walmart wasn’t making money being open overnight. Labor shortages and the resulting increase in labor costs have already trimmed profit margins, and it looks like Walmart pays a higher wage for overnight shifts. On top of that, I’m sure sales are lower at night.

I think most corporate decision making comes down to the question, “how do we max out profits?”

What’s Next?

Week in Review tomorrow (Friday), Part II of the January Q&A sometime next week, and our recap of the monthly jobs report next Friday.