[NOTE: 20 April, 2024. This started as an initial draft. I have made multiple updates to this post and will continue to.]
The middle-class in the U.S. is at risk. We often see information that shows the U.S. middle-class has greater wealth, including larger homes and more disposable income, than the middle-class in other western countries. But that is at risk of changing.
This raises the question: "What is the middle-class?" There are a number of definitions from a number of organizations. The Urban Institute defines it as 150% to 500% of the federal poverty level. This definition starts at the lowest level, and tops out at the lowest level. It is fair to say it includes the "lower middle-class" and excludes the "upper middle-class." The Brookings Institution used 20th to 80th percentiles of income. This starts at a point similar to the Urban Institute but goes much higher, and therefore includes what one might consider the "upper middle-class." Pew Research Center uses 67% to 200% of the national median income. This starts much higher than the Urban Institute or Brookings, but goes up to a similar point as Brookings, so it does not include the "lower middle-class" but does include the "upper middle-class." From these three definitions, we can roughly define a lower middle-class, a "middle middle-class", and and upper middle-class.
The first challenge for the middle-class is home ownership has become something only affordable to the upper middle class and wealthy.
The most recent data as of April 2024 shows the median monthly payment for a home purchase rose to $2,775, which requires an annual household income of about $120,000, which is roughly 1.5 times the median U.S. household income.
To understand this, the median monthly mortgage payment in 2021 was $1,001, according to Census housing data. This number represents the average of all mortgage payments, not just recent home purchases.
In 2021, the median monthly payment for a home purchase made that year was about $1,785, which required a household income of $76,500. How many people are making 57% more in 2024 vs. 2021? Housing prices really started shooting up in 2020. If you go back to 2016, the monthly payment for a home purchase that year would have been about $1,428 (requiring $61,200 annual income). If we go back to late 2011, not long after the historically low home prices coming out of the 2008-2009 collapse, a monthly payment for a home purchase that year would be about $1,140 (requiring $48,857 annual income).
The dramatic house inflation that spiked due to the decrease in inventory due to the pandemic caused construction halt, and now the high mortgage interest rates due to the current inflation, have combined to create a "double whammy." A doubling of costs over a decade, while the CPI has going up 32%, and wages have increased by 44%.
Here is something else to think about. If you follow Dave Ramsey's guidance, you should only buy a home on a 15-year fixed rate mortgage and not pay more than 25% of your take-home pay on total housing costs (mortgage principal and interest, home insurance, and property taxes). And make a 20% down payment. To be fair, when Ramsey talks about take home pay, he means gross pay less taxes and normal deductions like health insurance, but not 401k deductions.
What this comes to on average today, based on today's mortgage rates, is you should not buy a house priced more than twice your gross household income. Which means we could not afford the house we currently live in unless we made a 25% down payment, which would be $150,000. That is reasonable if one has a current home with significant equity they are selling, but not realistic if it is a first-time home purchase.
The area we live has gone from what was once probably a 75th percentile income region 20 years ago, to an 80th percentile income region 10 years ago, to a 90th percentile income region today.
To put that into perspective, you must earn 15% more to move from the 75th to the 80th income percentile, and then another 40% more to move from the 80th to 90th income percentile.
Unless something changes, in a generation, instead of three elementary schools within two miles driving distance from our house, there will be at most one, because families with young children will be priced out of the region. Instead of three middle schools and three high schools in the city there will be two, or perhaps just one. Prior to 1991, there were three elementary schools and no middle school or high school in this community. Students were bussed to middle and high schools to the west in the same county. In 1991 a middle school and a high school were built in the community. Rapid growth saw five elementary schools added between 1994 and 2000, another middle school added in 2001, another high school in 2002, another middle school in 2003, another elementary school in 2004, and another high school in 2008. From the way this growth was staggered, it is clear there were a lot of younger children in the community thirty years ago. The last school built in the community was a high school built in 2008. By this point the community had become solidly upper middle class, and many of the people moving into the community were middle aged people with high school aged children. That said, the local public school system is well attended by children of high earning residents living in very expensive homes. The middle school built in 2003 sits directly across the street from a very high-end country club. Directly across from the entrance of the school is the gated entrance and security guard building for this country club. And if you happen to drive by as school lets out, a steady stream of middle schoolers diligently walking across the crosswalk into the country club neighborhood. But yes, a 50 year-old corporate executive might have a youngest child who is a 6th grader. People are getting married later and having children later. But the other reality is the homes in this country club were in the $750 thousand to $1.5 million dollar range a decade ago, and now you probably will not find a home for less than $1.5 million in the neighborhood. It is a fact a decade ago a 40 year-old sales executive could afford a one million dollar country club home, especially given interest rates at the time. They would not be buying the smallest home in the country club, but a mid-tier home. But while the same sales executive today might still be able to afford a one million dollar home despite much higher interest rates because they earn more, the reality is the one million dollar homes cannot be found in the same country club, so they will need to step down to a less exclusive country club.
I remember a conversation relayed to me about my company's current worldwide executive vice president of sales. He bought into a large, expensive country club about ten miles north of where I live about a decade ago. This country club was famous because a former professional baseball player lived there. When someone made the comment the then lower-level sales VP must be doing good to be buying a home on that country club, he replied: "I have the least expensive home in the country club." Someone in the same role today he was in a decade ago probably could not buy into that country club today, even the least expensive home.
The same dynamic hits in lower income ranges and less expensive neighborhoods. It just keeps rolling down. The parents with young children will move further out into the exurbs, and suffer longer commutes, etc. Formerly solidly middle-class neighborhoods in the suburbs become exclusively upper middle-class.
A freezing in-place effect also occurs. The empty nester couple cannot afford to downsize because the higher interest rates make the less expensive condo's mortgage payment higher than the more expensive house they bought a decade ago.
Eventually inflation will come down, and interest rates will come down. As more housing is built, albeit in the exurbs, there will be downward pressure on housing costs. There might even be another housing price collapse like the one in 2008. All of this may correct some of the demographic changes.
However, the demographics of the area were already changing before the latest housing price increases and high interest rates. Registered students have been declining each year over the last several years for most schools. The decline is most noticeable at the elementary school level, but it is happening across the board. Given current housing prices, it is likely there will be demographic collapse of elementary aged children in the community over the next decade. This could be a temporary trough, or permanent.
The second reality is having more than two children with the intention of sending them all to college is increasingly becoming an upper-middle class or even upper-class idea.
Here is some back of envelope math to consider:
A four-year degree at in-state rates at a public university will cost about $100,000.
A four-year degree at out-of-state rates at a public university will cost about $180,000.
A four-year degree at a private university will cost between $250,000 to $365,000. Some private universities costs are on par with in-state public universities, while elite universities are north of $350,000 for four years.
The cost of a university education inflates at about 8% per year, compared to historic inflation of less than 3% per year.
From birth to college entry is about 18 years. So, you get to save for a child's college during a period 22 years, with money starting to be withdrawn four years prior to final maturity. Additionally, saving for college is like saving for retirement. You can invest more aggressively early, but need to invest more conservatively as you get close to the college years.
Over the last 20 years, the S&P 500 average rate of return has been 9.8% (with dividends reinvested).
The reality is, you would need to start with $100,000 invested at the birth of a child to have enough to pay for college when they enter college assuming a normal investment strategy. What this means is you will need to save about $8,000 per child per year in a 529 to completely pay for even an in-state public university education assuming no other assistance or source of funding (such as your child working during college).
Here is something else to consider. To to an out of state public university today costs more in real dollars than it cost to go to Harvard in 1985. As I note above, it will cost about $180,000 to send a kid to an out of state public university starting this fall. If Harvard's cost had grown only in step with CPI since 1985, it would cost about $170,000 to go to Harvard. This means the economic burden on sending a kid to a public university in a neighboring state in 2024 is higher than the economic burden to send a kid to Harvard in 1985. How many early GenXers know someone who went to Harvard in 1985? What kind of family did they come from? Probably very upper-middle class or more accurately, wealthy. The middle class are simply being priced out of college for their kids.
The current birth rate in the U.S. has collapsed to 1.62 births per woman and is lower the higher one goes up in income (and the likelihood of college attendance), so demographics is doing its part, but it portends a significant drop in potential college students in twenty years.
The idea of a commodity (education is simply information and the transmission of it) inflating at 4-5X the rate of CPI should be untenable in a modern society. It is unsustainable. Are people really going to pay $180,000 for an in-state, public university bachelor's degree entering college for the high school class of 2030? What about $325,000 for a public university in another state? Until we address this, we will face a higher education crisis of one form (unaffordability) or another (the shuttering of half of universities). It is already starting, with thirty colleges and universities shutting down in 2023.
Is there an answer? Yes. What is the answer? The answer is we must accelerate the adoption of the most deflationary tool we have access to, and that is technology. Technology is inherently deflationary. Second, we must accelerate the adoption of the most productivity increasing tool we have access to, and that is technology. Technology is inherently productivity enhancing. But there is a tremendous risk, and it lies with those who would "Pause AI." The simple fact is, in 2024, pausing AI is pausing technology. AI is a feature, not an industry. No new technology tool will be implemented without AI features. Those who would "regulate math" should be viewed with great skepticism, and a doggedly pursued using a "follow the money" methodology. Many who are proposing to regulate AI are simply developers of proprietary AI who wish to use regulatory capture to protect their markets.
Technology holds the promise to finally stop the maddening inflation in the areas where we see inflation at significant rates above the Consumer Price Index (CPI). For example, health care and higher education have inflated at a rate far above the CPI. One only need to look at textbook prices. We have a control group, other books, to look at. What is worse is most textbooks contain material that is in the public domain. Certainly new textbooks are needed for certain subjects. Computer science programming classes went from COBOL to C to Java and now Rust. But the content in English literature, history, introductory chemistry, biology, and physics texts do not change rapidly.
Education, at its core, is the transmission of existing information. Every other industry in the business of transmitting existing information has seen a collapse in costs. Witness the change in transmitting a document over the last 50 years. We went from postal mail, to overnight express packages, to the fax machine, to transmitting digital documents via email. Accelerating the transmission of information was also a huge productivity boost. And it was this measurable productivity boost in the 1990s when PCs, local area networks, wide area networks, and the internet created an economic boom that made the sleepy, wooded, once exurb with some scattered homes and a couple of new country clubs into the vibrant, middle-class suburb I now live in. My lifetime has witnessed this wealth explosion where middle class neighborhoods were once defined by 1,400 square foot three bedroom ranch homes and 1,800 square foot four bedroom split-level homes, to where middle class neighborhoods looked like the upper middle class neighborhoods of the 1970s.
Now we see the reverse effect. And we need to stop it. There are late Millennials and Gen Z who now have children of their own, and like the cohorts who came before them, want to live in a single-family home in a good, safe neighborhood in a good school district, without an hour-plus commute.
At the same time, while the trendy thing might be to follow the "one and done" philosophy and only have one child, or "two and through" and have two children, I know a lot of older Millennial families who have three and four children. When I run the numbers on college for my two kids, it deeply concerns me for them. The higher education establishment needs neutron bombs exploded over every campus. It needs a radical rethink. To be fair, for the elite universities, neutron bombs will not be enough. Break out the Castle Bravo bombs.
For many decades, the median universities imitated the elite universities. The "Harvard Case Method" became the norm for MBA programs around the world. Berkeley's David Patterson's "Computer Architecture" textbook was the standard for computer engineering taught in every engineering school. Education is simply the transmission of information. There is the information, the transmitter (the professor), the medium (the classroom at the university), and the receiver (the student). The only difference between MIT and State Tech is the quality of the human components. MIT gets the top one tenth of one percent of the high school students looking to study engineering, and has the top professors in the world. But it's the same laws of Newton.
I first noticed the potential of technology to accelerate education twenty years ago. I read an article about the U.S. Army's Command and General Staff College's distance learning program. The U.S. Army's Command and General Staff College is a master's level program for Army Majors and other service O-4 grade military officers. A small percentage of top performers attend in residence. The rest attend via distance learning while continuing their full-time jobs. In the late 1990s and early 2000s, some innovate faculty members for the distance learning program determined the Internet, and Internet streaming provided a means to host and deliver live stream or recently recorded content. They were able to get top guest lecturers such as Colin Powell and Newt Gingrich. The in-residence school was incensed, because they could not get this caliber of guest lecturers, who would have to fly in Fort Leavenworth Kansas to speak.
The next realization was around 2017 or 2018. I attend a non-denominational, multi-campus "megachurch." I have been attending long enough I have seen the church multi-campus video system evolve. The church is unique in that they purchased a used high resolution video camera from NASA and used it to video the pastor speaking on stage and simulcast it to another auditorium in the same building where it was projected onto a 16 by 28 foot screen on the center of the stage at a one to one scale. The image on the center screen was life-sized, and had a high enough resolution to fool people into thinking the pastor was actually on the stage. This center camera was fixed. The 11 by 19 foot side screens in the main service, used for close ups, videos, etc. were also simulcast to similar screens in the second auditorium, but these used standard television cameras. When the church expanded to a second campus, the remote campus presented the same sermon series on a one-week delay, leveraging recordings of the main campus service. The recordings of the two video streams (main center screen and side screens) were time synchronized and required the use of hard drives for recording. The second campus had an identical projector and screen setup. By the time the third campus was added, WAN links had become powerful enough live transmission of the side screens was possible. A few years later, live network transmission of both the side and center screens were possible for both remote campuses.
Then the Georgia Board of Regents, which oversees all of the state higher education system of Georgia, consolidated many of the various university and junior colleges. Adjacent to my church's main campus was a remote campus for Georgia State University (a four-year institution), and a two-year junior college. These two institutions were merged, and the two adjacent buildings were combined into a common campus. Driving past them every Sunday, I thought: "There is no reason they cannot use the same technology our church does to "beam" a class from the downtown campus into the local suburban campus. Just have a graduate assistant in the local classroom to handle roll, test proctoring, etc." I thought back to my time as a undergraduate and graduate student in 1989 and 1990 when my university started a distance learning program using VHS video tapes. A handful of classrooms were remodeled. The classrooms were soundproofed, and cameras were added. One camera was in the ceiling, and zoomed into a drawing pad which replaced the transparency projector. Instead, it was projected via television onto a screen in the room. The VHS tapes featured a split-screen of the professor and the overhead camera. Two feeds. Just like our church. The technology is there. You do not need the life-sized center screen. You just need one camera on the lecturer, and one video feed of the whiteboard or PowerPoint presentation. Simply web chat would allow students in remote classrooms to ask questions of the professor. High quality digital video cameras today are very inexpensive. Internet bandwidth is cheap. SD-WAN over the Internet means dedicated WAN links are no longer required.
Education was ripe for a rethinking well before COVID. We have Khan Academy and other concepts. When the pandemic hit, and people could not go to testing centers for IT certifications, individual human proctors watched testing students via the student's laptop cameras to make sure they were not cheating. Today AI could easily do this.
The reality is, the Berkeleys and the MITs and the Harvards and the Georgia Techs could easily triple or quadruple their enrollment overnight if they wanted to. We could cut the number of professors needed in half tomorrow. We don't even need grad students as teaching assistants and proctors. Upper class undergraduates could do it for freshmen and sophomores, and master's students for upper classmen. We could have done this a decade ago.
Then there is the insane expansion of administrative positions in higher education. While every other industry has seen a collapse in overhead positions (just try to find a typing pool in the headquarters of a Fortune 500 company, or a secretary supporting a mid-level manager), higher education has seen an explosion of administrative overhead. In some cases, administrators outnumber academics at universities by three to one. AI could easily replace many administrative functions in higher education.
The ultimate goal should be to cut the cost of a bachelor's degree in half in ten years. That is a very reasonable goal. But it will probably also require a complete rethinking of the over a century university accrediting system. Accreditation is a form of regulation, and regulation is always an impediment to innovation. An obvious approach would be to, instead of accrediting universities, to accredit individual courses and curricula. I recall when my university was threatened with losing its accreditation due to violations in the athletic department (which should be the purview of the NCAA, not the accrediting association), and a loss of confidence in the university president at the time by some in the faculty (the university president being an administrative role, and not an academic role). This terrified the students, who were told if accreditation was lost, their degrees would be permanently worthless. What, pray tell, does the football team, or a university president getting on the wrong side of some professors, have to do with Aerospace Engineering?
Given the speed at which technology changes, and the requirement for lifelong learning, less formal education, more apprenticeship, and more on the job training make much more sense in the 21st century. One can argue the western university model is a millennium old. At a minimum, using the the original Morrill Land-Grant act, passed in 1862, as the start of the modern U.S. university system, our university model is over a century and a half old, roughly in line with the Second Industrial Revolution.
The time is ripe for a new experiment in post-secondary education. And that will require slaughtering some sacred cows.
AI and other technology will not build more housing, yet. But it can optimize traffic patterns, making more distant suburbs more valuable. We already go to the store less. I needed an Ethernet cable and got in my car to drive 25 minutes to Micro Center to buy one. Then, sitting at a traffic light I just ordered one on Amazon from my phone, and instead of turning left, I did a U-turn and went home.
What AI and other technology can do is make the upper middle class more productive and raise them to the upper class. What AI and other technology can do is make middle class more productive and raise them to the upper middle class. What AI and other technology can do is make the lower middle class more productive and raise them to the middle class. And it can take the working class and raise them to the lower-middle class. And it can do this while simultaneously lowering the costs of health care and higher education. It is a win-win-win-win-win-win situation. We are not taking jobs from anyone. With a 1.62 fertility rate everyone in the future is going to have a job. We might create so many jobs the federal government starts paying middle class people to have more kids. Imagine if the federal government offered free higher education to any second and third child born to a couple.
Acceleration of technology will simultaneously increase productivity and drive down costs. Accelerating technology is deflationary. Accelerating technology boosts productivity. This has been true since the discovery of fire and the invention of the wheel.
We must accelerate.
Effective accelerationism is the only philosophy that hold a reasonable promise of returning the U.S. to an economic curve where the middle class thrive.
Effective accelerationism is not something just for the Silicon Valley tech lords or the venture capitalists. It is not just for software and computers. It is also for manufacturing and reshoring. It is for nursing and elder care. It is for teaching and pediatricians.
We must accelerate. Or the middle class will die. Home ownership will be reserved for the wealthy, the vast majority will rent their homes from Blackrock, the federal government nationalizes the failed public university system, and dictates who gets to attend. If you think university admission is political now, just wait until you have to contribute to your local U.S. representatives re-election campaign to buy your child a chance to attend the middling state technical university.
No comments:
Post a Comment