Why are some industries fragmented while others are consolidated?

Great Debates
1

For example, trucking in the US is an incredibly fragmented industry, where the 50 largest companies handle just 30% of the volume and there are over 110,000 carriers and 350,000 owner operators. On the other hand, shipping is a pretty consolidated industry, where the top 4 shippers alone carry almost 50% of the volume. It’s not immediately obvious from the outside why trucking would be so fragmented while shipping would be so consolidated.

As another example, in Australia, there are many fast food restaurant chains but it’s hard to think of many examples of mid-range, family style restaurant chains. This is in contrast to America where you have dozens of well known chains like Olive Garden & Ruby Tuesday. It’s again not immediately obvious why such a business model would work in the US but not Australia?

Of course, for many industries, the reasons for consolidation are obvious. There’s a reason there’s only 2 major mobile operating systems or 2 makers of CPUs or 2 makers of long haul aircrafts, for example.

There’s also a reason for why things like restaurants will forever remain fragmented as each chef has a unique menu and quality of food that allows for differentiation. However, it’s not immediately obvious, for example, why so many laundromats are independently operated and there doesn’t really seem to exist any large, nationwide laundromat chain. People are mostly seeking the exact same thing from laundromats so why has the industry not consolidated?

It seems very interesting to me that certain industries that seem almost identical have very different structures. Is it the result of some historical accident or are there hidden forces I’m not aware of that keep certain industries the way they are?

2

Think about the costs of hauling a truckload of merchandise from point A to point B: the small company (maybe only one person) and the large national firm. Both will use the same amount of gas and both will take the same amount of labor for the driver. There is not much in the way of economies of scale (The big company can probably buy trucks and other stuff a little cheaper. On the other hand the big company has a big bureaucracy to pay for…)

Think about the shipping company picking up and delivering parcels. The small company picks or delivers 10 packages in a trip in a distance X miles long. The big company has a lot more business and picks up or delivers 100 packages in this X mile distance. On a per package basis the big company has a lot lower cost. Thus there are very few package delivery companies.

3

I think a big trucking company would have the same benefits a package delivery company would have: a larger pool of contracts. You might not be able to deliver any individual contract cheaper but you’d be better placed for “synergy” (I know that’s a business cliche but it’s applicable here).

A big company with thousands of contracts is going to have times when several contracts are all traveling on the same route and can be combined. Or when a delivery from City A to City B matches up with a delivery from City B to City A. A small company with only dozens of contracts will end up delivering each one individually.

4

I am talking about truckload transportation. When multiple loads can be combined that is called LTL (less than truckload) transportation. That industry is much more consolidated than the truckload segment, but less consolidated than the package delivery companies. Here are the top LTL 25 carriers:

5

The package delivery business requires the operator to set up a system to deliver to every address. That’s expensive, although if you contract with the USPS for last-mile delivery, it might be feasible to introduce a competitor.

Sometimes, it’s a matter of a new operator consolidating an industry. Wayne Huizenga did this, first in the garbage business with Waste Management Inc and then in the video rental business with Blockbuster Video. In both cases, the industry was previously very fragmented.

Today, it appears that Mattress Firm is doing the same thing with mattress retailing.

6

I wonder about this sort of thing a lot. I see trucks with 80,000 pound gross vehicle weight stuffed to the gunwales with maybe 8,000 pounds of potato chips, and right next to it a truck only 20% full by volume but at the weight limit with machine tools. It seemed like with gps and computer routing and load matching there would be a lot of synergy in putting these loads together. Get a load 80% full of something light and bulky matched with a high density shipment.

7

I’m less interested in the specific factors that affect a particular industry and this more about the meta-structure across all industry. Like, does there exist a relatively simple set of rules such that, given a certain industry, it’s possible to predict how fragmented it is or is the degree of fragmentation a result of historical accident and chance?

8

I’m not an expert, just an observer, but I think the answer to this question is a firm no.

Not because maths wouldn’t apply to a given industry in general, but because it isn’t JUST the actions involved with the particular industry itself, that determines costs and business structures.

One partial example: a lot of industries that USED to be more centralized, with a few large operators, became much more fractured, because tax and other regulatory laws were passed, which accidentally favored subcontracting over directly employing people.

Accounting philosophies have affected this as well. When I was very young, most large companies counted costs and profits in a different way than they do now. Elements such as parts was considered a single cost, and profits were calculated on a large scale for the entire company. Then in the 70’s, a new concept came into vogue, where large companies were artificially broken down into semi-autonomous units, which they called “profit centers,” to make the idea sound less stupid than it was. Suddenly the parts department had to show a profit by itself, so it could no longer “sell” it’s parts to Manufacturing at “cost.” Prices rose accordingly, and taxes went down, because the internal “costs” of subdivision profits, could be deducted from overall “profits.”

9

Coase wrote the definitive treatise back in the 1930s. The Nature of the Firm - Wikipedia The optimal size of a firm is driven by the balance of communications and coordination costs versus capital requirements.

The implication is that simple businesses could be small and not vertically integrated. Complex businesses needed vertical integration to bring the required coordination in-house.

Technology gets a vote too. What Coase said was true then and still is true now … for any given level of tech. Modern communications and internet services has relatively reduced the cost of coordination. And therefore reduced the threshold above which businesses must integrate vertically. Apple today doesn’t need to be as integrated as Ford (or IBM) was in the 1930s or 1950s.

So that’s vertical integration.

Which is separate from, but not totally irrelevant to, horizontal integration. AKA industry-wide consolidation or fragmentation.

As a general matter industries consolidate if there are positive returns to scale. Net of, as **igor **just said, externalities like regulation or taxes.

When interest rates are low, borrowing capital to hoover up competitors becomes easy. When rates are high it’s tough. So an industry which does well in a low interest rate environment will be more consolidateable than one that’s on death’s door each time the economic cycle delivers low interest rates.
There also is a lot of path dependency to history. Beta might have won over VHS. The US might have chosen to drive on the left side of the road. Some industries are consolidated now because they happened to pass through a propitious environment 10 or 50 or 100 years ago. Others not.

The airlines have consolidated because network effects predominate for a transport service running on a schedule with multiple independent purchasers per vehicle trip. The airline that can offer 500 city pairs does more than twice the business and enjoys better margins than one who can only offer 250 city pairs. OTOH trucking has not consolidated because each truckload is custom scheduled for a single customer. (Less LTLs mentioned above).

Tech like Uber raises the interesting question. Imagine an Uber-like service for the long-haul trucking industry. The booking advantage the larger truck lines have over the independents would all but evaporate. Their cost disadvantage would remain. We see that tech may change the answer to the question of whether there are returns to scale or not.

But the advent of Uber-like businesses raises a different meta-question: Whether a hypothetical Uber-ized trucking industry should be thought of as one highly consolidated industry under Uber HQ or a vastly more fragmented industry of individual sub-contractors?

I would answer the question by asking and answering another: In that environment if Uber itself disappeared tomorrow, would the systemic industry-level disruption be large or small? If disruption is large, then Uber itself is the relevant unit of measure and it’s highly consolidated. If disruption is small, then the industry is unconsolidated.
Lots of different aspects to this stuff.

10

Its usually some combination of barriers to entry, economies of scale, regulation, risk reward profiles.

11

Well, think of the cost of (1) one truck vs the cost of (1) Maersk Triple E class container ship, plus at least two intermodal container ports to load and unload them. Plus the containers ship (by definition) doesn’t care if 10,000 trucks each from a different company drops off a container.

Generally speaking, the more capital and resources it takes to start a business, the more likely economies of scale will encourage businesses in that industry to consolidate.