Americans have the illusion of choice, but in industry after industry, a few players dominate the entire market:
- Two corporations control 90% of the beer Americans drink.
- When it comes to high-speed internet access, almost all markets are local monopolies; over 75 percent of households have no choice with only one provider.
- Four airlines completely dominate airline traffic, often enjoying local monopolies or duopolies in their regional hubs. Five banks control about half of the nation’s banking assets.
- Many states have health insurance markets where the top two insurers have 80-90% market share. For example, in Alabama one company has 84% market share and in Hawaii one has 65% market share.
- Four players control the entire US beef market.
- After two mergers this year, three companies will control 70 percent of the world’s pesticide market and 80 percent of the US corn-seed market.
The list of industries with dominant players is endless.
Pointer from John Mauldin. Read the whole thing. Tepper also makes the case that many local labor markets are monopsonistic, meaning that only a few employers are available.
A recent paper by James Traina in the Working Paper series of the Univ. of Chicago’s Stigler Center (February 2018) shows that markups, when properly measured, have not increased over the last few decades. It would appear that, even if many markets seem to be dominated by only a few players, these large firms must be constrained in their ability to raise prices, perhaps by the threat of entry by new technologies (what Baumol labeled “contestable markets”).
https://promarket.org/are-markups-increasing/
Tepper’s argument aside, I wonder what the worker compensation timeseries would look like if benefits (specifically employer-paid healthcare costs) were included.
And taxes
Mr. Tepper refers to his article as “An Economic Detective Story”. It strikes me as nothing more than a poorly considered review of the “Usual Suspects”.
And I’m thinking here of one very blatant mis-representation of “wage-earners” and one very glaring omission in his calculus of “wages”.
The blatant mis-representation is the implication that “wage earners” are not only a different group than corporate owners/stockholders, but they are indeed mutually exclusive groups. Nothing could be further from the truth, and even the implication is insulting. Where is his “detecting” that the vast majority of “wage earners” are either directly owners of stock in these firms, or indirectly owners via their pension and retirement accounts?
The glaring omission in his calculus of “wages” is that of health care and health insurance costs, not to mention employer contributions to retirement/pension funding. Perhaps a bit more “detective” work by Tepper might reveal the fact that very close to 20% of GDP is health care and health insurance costs – ALL diverted from “wage earners” and “wages”.
The focus on wages paid and prices charged seems less important than the fact of the very widespread trend in corporate centralization, with whole sectors consolidating or being dominated by one or a few “big winners.” The explanation behind that is probably driving the other things we observe.
My explanation is that Thiel-Porter theory (for lack of a better name) is true. That is, there is a kind of natural selection at work, and corporations – especially those in expensive developed countries – without some special insulation from upstart competition, or benefiting from barriers to entry, will not be able to stay strongly profitable, because some cheaper copycat abroad will be able to arbitrage and eat their lunch.
So, the survivors will all have a special something. A good and classic candidate for one of those special somethings is “positive economies of scale / scope” which includes matters related to “network effects.”
But that possibility by itself doesn’t lead to widespread centralization and consolidation, because “diseconomies of scale and scope” could also be a special something that would explain small, local survivors. Indeed, the need for Hayek’s “local knowledge”, if only readily available to and exploitable by a local agent or arbitrageur, would provide precisely that kind of diseconomy.
But we’re seeing tne economy evolving in the direction of most output coming from giant global behemoth winners and not small, regional survivors able to leverage their superior command of local knowledge.
As I explained here in a previous comment, that’s because something big has changed recently: locals are no longer best at local knowledge. The ultimate form of local knowledge (or “market research”) is conducting comprehensive electronic surveillance on everything every individual does, to be able to discover and infer all kinds of things about the economically-relevent aspects of their personality and psychology, that is, “economically actionable intelligence.”
And it turns out this kind of economic intelligence collection can only be done by a few large entities and has huge positive economies of scale since marginal costs are extremely low.
So, since the survivors of corporate natural selection with be those with a special competition-insulating advantage, positive economies of scale provide such an advantage, and even “command of local knowledge” now has flipped from negative to positive economies of scale, we should expect widespread centralization and consoliation and most sectors to come to be dominated by one or a few giants.
Interesting. How long do you expect this trend to continue, and do you see any countervailing forces?
It seems possible that some combination privacy laws restricting data collection/retention on the one hand, and laws mandating the publication of data on the other, could undercut much of the advantage of these monoliths. However, I have little hope that any such legislation could pass without being hobbled to the point of uselessness. Furthermore, I’m sure that many governments are content to allow the accumulation of corporate surveillance so long as they can demand special access to the contents of that data.
Eric Raymond (of The Cathedral and the Bazaar fame) also makes interesting case regarding “secrecy rent”. Essentially, those companies that primarily capitalize on extracting rent on secrets reach a point that they can no longer provide value-add from those secrets. Thus, in order for those companies to maintain the growth demanded by their shareholders, they must exploit their customers to the point where the customers are willing to absorb the cost of transitioning to alternatives.
Now, his argument was mostly regarding secrecy of software algorithms, but it seems plausible that it could apply to data as well.
The “special something” could be derived secrecy, or economies of scale, or a variety of other things.
But the modern global economy is very much a “Get Rents or Die” environment.
I gotta disagree with the second part. The “economically actionable intelligence” is to me still pretty darn crude. Google Maps keeps asking me to rate my visit to the same effing stores I go to every week. Google knows where I live, it knows what I’m planning to purchase, via my Chrome browser history, and it knows a lot of my interests, too, via location tracking, like when I go golfing or to the park to play tennis, and they know who my friends and family are via my gmail address book. And yet, what are they able to do with this information? Not much, from what I can tell. The next time they’re able to actually influence a purchase of mine will be the first. Are they even trying? At best I get little ads at the top of my gmail tabs or a search results page, which I ignore. Maybe they sell this information and I get fliers I don’t read in the mail?
In addition, we should consider the “social norms” multiplier effect on networks.
By this I mean, not doing things because of connections/compatibility (network effects) but doing things (buying particular products) because your community, peer group, etc. do. Or using a particular service out of habit (see amazon prime…)
I suspect that for example Apple largely survives on this basis, that is people buy iphones because they bought iphones before. There are still a few apps that are iphone only, but they are quite rare.
This is a rational thing to do (at least up to a point) because costs in change of habit pattern (let alone network costs) of switching are high, and the return generally low.
Changing from selling a thing to extracting a kind of never ending rent for it while not necessarily adding any value is obviously a big thing in computer software, but also in odd places like parts/maintence on agricultural equipment.
I think a key trick to maintaining such rents is that most consumers (persons or firms) perceive the rents to be low or at least tolerable.
Reading Tepper’s post, I ran into several issues with the facts he uses in his analysis. Iwill just mention a couple of the largest that I saw.
First, Tepper brings up the divergence between productivity and hourly wages. In the chart he shows us, he show wages and productivity diverging since the 1970s. He also brings some some issues that are brought up with this chart and then shoots them down. The problem is he does not bring up the main objection that people have to this chart; namely the use of two different adjustment factors for inflation. Wages are adjusted using the CPI while productivity uses the GDP deflator. The divergence decreases significantly when the same inflation adjustment factor is used.
Second, Tepper mentions that the number of publicly listed companies has declined significantly over the past 20 years. Becuase of this Tepper argues that the number of companies has been deciling while the economy grows. This would indicate that businesses are becoming more concentrated. I think this change is not due to a decrease in the number of companies, but in how they are organized. If you look at corporate tax returns (https://www.irs.gov/statistics/soi-tax-stats-integrated-business-data) you do see that the number of C-corporations in the US has declined from more than 2.5 million in 1986 to 1.6 million in 2013. However, this decrease is due to a shift from C-corporations to passthrough entities. While the number of c-corps decreased by nearly 1 million, the number of pass-through entities increased by nearly 5.8 million. Overall, the number of corporations filing a return increased from 13 million in 1980 to 33 million in 2013. How does Tepper address this issue? He does not in his article. Perhaps this will be in the book he is working on?
Third, some already commented on the issue that there is less agreement on markups than Tepper addresses in his article. Other research has found a lot smaller markups than the source Tepper uses. Who is correct? I don’t know, but again Tepper does not even address this.
Fourth, Tepper mentions that CEO pay has increased signficnatly faster than worker pay. I believe Mark Perry at his Carpe Diem blog has talked about this issue in the past. Teppers comparison looks at CEO pay of the largest 500 corporations and then compares it to the compensation of all workers. If you look at all CEOs, or only the workers who work at the largest 500 companies, the change in CEO pay becomes significantly smaller.
Overall, I have heard Tepper’s arguments before from different sources. While there is some useful information and things that need to be looked at in more detail, I think the strength of Tepper’s arguements are far weaker than he presents.
No price elasticity in labor mean workers get bottlenecked.
Wage elasticity is gone because 30% of wages are fixed fees for governments services. Hiring managers are competing against those fees and need to be as big as government.
Several of his illustrations regarding market concentration, if not the majority of them, are that way largely due to government interference in the market. I see no evidence in Tepper’s working paper (although I only skimmed it briefly) that it even occurred to him to look at private vs. public “firms” — although I may have spent enough formative years in small university towns to prejudice me, I suspect it also may be true that government employment skews the results here. (If I am wrong about Tepper’s working paper, please correct me.)
+1
Came to leave substantially the same comment. The best examples of concentration appear to be driven by regulatory burdens driving smaller companies out of business, or preventing them from forming, or limiting competition.
It seems to me what this shows is that having an industry dominated by a small number of companies does not in itself harm consumers. The way I judge whether consumers of X (some product or service) are being harmed is whether they can only obtain X by submitting to conditions of contract that many would consider outrageous. Thus in my judgment beer consumers are not being harmed, but consumers of banking, insurance, health care, rental housing, and computer operating systems are.
In most of these cases overregulation caused the harm (and also caused the small number of competitors in the industry). That includes rental housing if you count zoning as regulation (because it keeps potential competitors out). But in the last case we have to wonder what is keeping new competition away when so many consumers are dissatisfied with what they’re getting now.
There are three exceptions I take with this type of approach. First is that there is no consistent way used to classify market dominance. If you compare the beer example to the bank example the banks seem positively diverse! 90% of all beer controlled by 2 companies vs 50% of all assets controlled by 5 banks, that makes the the largest brewers 4-5x as large (proportionate to their market size) as the average large bank.
Secondly it cherry picks numbers. Two large breweries produce 90% of the beer drunk BY VOLUME, but that number drops to under 80% by dollar value of market share. Additionally this number has been declining for 20+ years now, the number of small breweries has more than tripled sin 2009, with steady growth in volume, dollar sales, jobs etc over that stretch in both absolute and relative terms. Finally beer is far from the only alcoholic beverage, let alone any kind of beverage, on the market making the distinction on its own pretty meaningless.
These complaints extend to most of the list given. 3 companies control 70% of the corn seed market, is there something specific about corn seed that makes it dangerous or undesirable to have this situation occur? That is one section of the entire corn market which includes growers, shippers and refiners of corn which is itself one section of the overall food market.
I’m very confused, most likely due to ignorance (demonstrated below).
I keep hearing this [high number]% of beer Americans drink is controlled by just [small number] of companies. The numbers change (see here for 70% by 2 companies https://www.foodandwaterwatch.org/impact/70-percent-us-beer-controlled-two-companies-want-merge), but the message is the same: We face fake choice when it comes to beer.
First, source please.
Second, this is super imprecise. What does “control” mean? Does it mean 1) these 2-3 companies have shares in the seemingly endless little breweries that are popping up all over, 2) these companies have controlling shares in these little breweries, or 3) that they CONTROL what ingredients go into each beer?
If it’s anything other than the last one (controlling the ingredients), who cares? Every town I go to has some local beer that I try. It’s sufficiently different (to me) from my hometown’s many microbrews that it gives me something interesting for my $5, and it’s within the range of what blonde, kolsch, pilsner etc. means that I know what I’m getting. Who cares in InBev owns controlling shares in my hometown’s microbrew and in town x’s microbrew? It’s a different experience than getting the same Bud, Miller, Molson, or Guinness all over the world (something I fall back on when the beer list gets too complicated).
Am I missing something? I doubt that InBev is telling TownXLocalBrewWithQuirkyName what ingredients to put in their beer recipe. Why is it a big deal that big companies that are good at making, marketing, and distributing beer have, at most, controlling shares in smaller beer companies?
Choice from this consumer’s perspective depends on what I’m tasting/experiencing, not what company my money ultimately trickles up to.
It doesn’t mean “control” at all. It’s a misleading locution which means nothing more than “By volume, 90% of the beer drunk in the United States is sold by two companies.” It would be just as accurate to say, “By offering superior value, two companies have won 90% of the domestic beer market.”
One concern I have with the tone of Tepper’s paper is the idea that increased productivity should drive up the price of labor. Increased productivity allows companies to pay more for labor, and increased profits create incentives to pay more for labor, but labor prices are not going to go up if there is a significant increase in the supply of labor. Over the last 50 years many more women went into jobs previously reserved for men. This increase in the supply of labor is one factor that could account for the lack of wage growth.
Tepper writes, “While many executives go on the front cover of Fortune or Forbes and get all the credit for their company stock, worker productivity has been steadily rising for decades. Unfortunately, earnings have not kept up with productivity increases. Workers are producing more goods with less labor, and companies are making higher profits, but the benefits are not being shared with workers.” One of us doesn’t understand productivity. Since he is obviously much smarter than I am, I wish he would explain how worker productivity increased. I suspect it went up because the managers he thinks aren’t responsible for the productivity increase were actually the ones who raised the capital to invest in the equipment and technology that increased the productivity. He appears to think that worker productivity has gone up because of something the workers did.
Maybe he’s right, but he didn’t present any evidence to that effect.
“The list of industries with dominant players is endless.” << and much management theory is telling managers they should be 1 or 2 in any market segment they are in, or they shouldn't be there at all.
Globalization means even if there was only 1 US corn-seed maker, they couldn't raise their prices much above that of Canada's or Europe's top corn seed supplier.
The more that stuff bought is a "commodity", like Bud or Miller, the tougher it is to increase margins. And more of what we're buying is being made as a commodity, including the push for mass customization.
Wages won't go up until labor is a bigger constraint than market share — and those who "want" to hire new workers can also, instead, raise their prices a bit.