Competition inside corporations?

Thursday 18 September 2008

Having observed Hewlett-Packard from the inside for almost 18 months now, I'm struck by a paradox: our economy is a chaotic marketplace of capitalist competition, practiced and championed by corporations, but internally, companies are run as top-down, centrally-planned dictatorships. Why is that? Why isn't a company simply a microcosm of the larger economy?

Take the case of IT services: inside HP, there is a large IT organization, and they provide services to the rest of the company. When my group joined HP, we had no choice about how to get, for example, email service. The IT group provided email, and we used it. When we need to buy a laptop, there is one group that provides that service. When we need servers hosted, we have only one place to turn.

I'm sure the reason for this is the efficiency gained by eliminating redundancy. If there were two groups providing email services, surely one group could do the job of both, with less total staff, equipment, and so on.

That's certainly true, but then why don't we apply the same logic to the larger economy? After all, HP's email group has a huge overlap with Dell's, IBM's, Sun's, Microsoft's, and so on. Couldn't our economy gain by eliminating the overlap? When these questions are considered at the national level, we tout the increased efficiency produced by competition. The economy as a whole gains from the pressure competition puts on each company. Without competition, there is no incentive to improve, no reason to do your best. In a centrally-planned nationalized economy, incompetence is not punished, incentives are mis-aligned, and apathy takes over. There's no reason to improve because your customers have nowhere else to turn, poor service will not lead to loss of business, there's no price pressure, and your existence is guaranteed by the state.

That's logic that every capitalist believes, and we laugh at economies that have tried central planning and failed. So why doesn't the same logic hold inside companies? Why are monopolies and lack of competition not just accepted, but enforced? Don't we believe the same forces will be at work? Is there any compelling reason to improve if you have no competition?

Why couldn't a company have three IT groups (call them Red, Green, and Blue). Each is separate, and lives or dies based on their ability to attract business from the rest of the company. When my group needs servers hosted, we shop around. Maybe Red is the deluxe service, and Blue is economy, and we've heard from friends that Green has the best service. For whatever reason, we choose one of them, and spend our internal dollars with them. The groups will compete, and that competition will force them to optimize and find the best solutions for their customers. If they don't, they will go out of business.

I know it seems wasteful to have all that going on inside a company. There will be duplication. But remember the capitalist logic: without competition, there's no reason to do your best. Just as with the larger economy, the duplication will be worth it because of the increased efficiency forced by competition. And without competition, your only option will be a poor one.

Of course, not all work inside corporations could be run this way. For example, legal departments deal with the outside world, and the corporation must speak with one voice there. But couldn't competition be used in at least some parts of large companies?

Where's the flaw in this logic? Why isn't competition inside corporations a good idea?


James Tauber 10:04 PM on 18 Sep 2008

Paul Graham wrote a bit about this a few years ago in an excellent piece I quoted from here.

I think one thing to consider is, if HP, or another large company, did consist of competing Red Green and Blue teams, how would that be different that three separate companies? Which leads one to think about Coase's Nobel-prize-winning work on the nature of the firm and why we have firms at all, given if we follow your thinking to its logical conclusion we'd all be freelancers working for ourselves.

The reason Coase gave was that there are certain transactional costs in the market: finding who to trade with, negotiating the terms of the trade and actually executing the trade that in *some* cases are lowered within a firm. To the extend those costs can be lowered, the need for firms is reduced.

A great example (I think from Don Tapscott) of how things have changed as certain transactional costs have lowered, consider that the big car companies used to own the shipping lines that transported their materials and owned the rubber plantations to make the tyres for their cars, etc. At some point, the transaction costs dropped enough that it didn't make sense for the car manufacturers to do that internally.

Robin Munn 10:47 PM on 18 Sep 2008

The flaw in the logic seems glaringly obvious to me. The difference between a company and the economy in general is that a company is a single economic unit. It takes in money through sales, or service contracts with the outside world. Internal "contracts" -- e.g., between the sales team and the I.T. team -- bring in no revenue for the company. Even if you run your company's departments like an economy, where the sales department pays the I.T. department for I.T.'s services, there is no increase in the company's bottom line as a whole. Which means that multiple I.T. departments are a total loss for the company. The company's income has to pay the salaries of all those people, but the extra people are bringing in zero net dollars to offset the cost of their salaries.

The competition aspect would be beneficial, yes, but it's completely outweighed by the fact that the company has to pay the salaries of all those people while they compete with each other. In an economy, the three competing companies would not have financial interdependence; each one pays their staff's salary with the money they take in, but there's no single revenue source that has to pay for all three. Therefore, the fact of having three companies doing the same thing isn't a resource drain, and the time it takes for one to go under (or for all three to establish themselves firmly in the market with separate customer bases). Whereas three I.T. departments within the same company are all draining the same limited resource -- company income, and until one or two of them go under the company is paying a LOT of salaries and getting no net benefit from all the extra employees. You might end up with a more efficient I.T. department in the end, but the amount of money invested in the process wouldn't be worth it.

Am I making any sense here? I know you're a smart guy, and this seems so glaringly obvious to me that I can't believe you missed it. But I don't see you address this at all in your article, so maybe the article needs to be expanded?

Jack Diederich 10:56 PM on 18 Sep 2008

The answer is signals. In a company the signal is dollars. A division of HP will be killed if it gets too expensive because it is easy to see how much it costs relative to a 3rd party vendor or a similar division at another company.

The same is not true for government where the signal is votes. Votes are very abstract things. You vote for a guy who has a slate of policy promises (that you like slightly more than the other guy) and then he votes for a policy and then somewhere down the road it is implemented by a government employee who you never voted for at all.

Put differently: politicians win elections by being the best at winning elections. Part of their appeal is the appearance of job competence but a lot of it is who they hire to do TV-ads. Faking a job interview is possible too but in private practice those people get weeded out more quickly because their job is more concrete and they are judged by people familiar with their supposed competency.

If companies were run like the government the Computer Engineers Union at HP could lobby the company to give pay raises to Engineers. Needless to say they would spend more time on PR and less on actual work than they would have otherwise.

[for more stuff on the very inaccurate nature of votes as signals see Von Mises.]

Jack Diederich 11:00 PM on 18 Sep 2008

Aside: there are actually Unionized software people. Many years ago I interviewed at the Wall Street Journal and I was told I had to join the Typesetters Union if I wanted the job. I declined.

James Tauber 11:07 PM on 18 Sep 2008

I'm not actually sure I agree with Robin Munn's argument above. For example the statement "The competition aspect would be beneficial, yes, but it's completely outweighed by the fact that the company has to pay the salaries of all those people while they compete with each other." doesn't seem necessarily true.

Consider: would it make sense for a VC to invest in two competing companies? I would think, purely economically, it might.

Jason 11:39 PM on 18 Sep 2008

I'd imagine the people working there wouldn't be particularly happy competing with each other, and even more so if everyone is at the same campus. Office politics is already bad enough when departments are implicitly competing for funding, but if you raise the stakes it might get ugly enough that your engineers all leave for greener pastures.

As others pointed out, accounting can be tricky. Unless you have everything decentralized and partitioned so that each division is funded solely by the revenue that its products bring in -- at which point they're practically their own company -- you need to have higher management decide your corporate strategy. Maybe low-end products are going to be the next big money-maker, so the company decides to invest more in developing those products. If you have strong competition (with real penalties for underperforming) between divisions, people who aren't developing the low-end products are going to be resentful, and it's not like they can just come up with a new deluxe product that will redefine the market. They might even lobby their higher ups against working on a new, potentially revolutionary product, to protect their own budget.

None of this is good for the company, especially in the long term. Of course, there's probably a good in-between.

Doug Napoleone 11:51 PM on 18 Sep 2008

Many companies do have competition within and between individual business units and teams. There is competition at some level within Google and Microsoft. And there is competition between groups is smaller companies as well. Some of it rises from skunk works projects, some of it occurs due to having duplication in technology due to purchases and mergers. 'Synergy' is rarely a clear cut, simple process, and is more often a drag out battle between groups.

One thing which can help foster this type of competition is by breaking up a company into business units where time and technology are billed between the units. Corporations seem to love this system to help them identify where the money is being spent, and which units are not actually profitable. Motorola used to love this type of accounting back when I worked there.

Would like to give more concrete examples, but I am living in interesting times....

Rami 3:59 AM on 19 Sep 2008

There's a whole range of reasons why comparing a company to a national economy simply falls apart -- for one thing, human rights come into play (modern perception is that a company can, perfectly fairly, remove a suboptimally performing employee by firing them; a nation cannot in good conscience deport or execute a suboptimally performing citizen). But to work within your example, I think the question is one of control. An internal company solution is less efficient cost-wise but preferable because at that decision point, control over the service (e.g. ability to put in place policies on employees' email) is considered more important than pure cost. In a perfectly competitive world, utility functions like email and server hosting would be outsourced (as other commenters have suggested) and it would be possible that IBM, Sun and Dell all used GlobalEmail Inc for their email needs and GlobalServerSolutions for their hosting. But that would deprive them of control...

Jeff Darcy 6:01 AM on 19 Sep 2008

Having three IT teams might ensure that one of them was better than it would be otherwise, but you'd still be stuck paying the costs of the other two because the fixed costs of employing those people would still apply even if they're idle. If you let them "go out of business" in that case then you're back to monopoly. Throw in the transaction-cost issue, and the fact that the most "profitable" IT tasks might not be the most necessary, and you'd have a mess on your hands. In the broader market, the risk/cost of being inferior or misguided is borne by the failing company; within a company, that risk/cost would be borne by the company and its shareholders.

IT is probably a bad example, though. I think internal competition can work better in product development, when it's used as a hedge against uncertainty. Sometimes you just don't know which architecture will work better, or which product variant will be more appealing, until you bring both to a stage of significant (though less than product-level) completion. In those cases, it makes sense to set up some internal competition. When you know what you're making and how you'll make it, though, it makes less sense.

James: VCs invest in competing companies all the time. In fact, since many VCs tend to specialize in certain segments, it's pretty much unavoidable if they want to put their own investors' dollars to work. The difference is that VCs are speculative by nature. Product- or service-oriented companies aren't supposed to be that way. Consistently spending shareholder money on projects that each have a 2/3 failure rate would be considered irresponsible and a breach of fiduciary duty, just as spending that money on pork-belly futures would be.

Paul Downs 9:04 AM on 19 Sep 2008

Also: who would set up the three competing IT divisions? It would be like picking teams in the schoolyard. If you give all the stars to one team, then why have the other teams? If you distribute talent equally amongst the three teams, you are lowering the capabilities of each of them. One thing that large companies have which the economy as a whole lacks is a layer of management, interested in the overall profitability of the company, which can evaluate the performance of each division within the company. It is that management group's responsibility to make sure that each unit of the company performs. In this day and age they have the option to outsource any function of the company at any time. If your IT people are smart they realize this, and so there are at least two corrective incentives to poor performance: fear of the boss, and fear of having your job outsourced.

Nicola Larosa 9:06 AM on 19 Sep 2008

John Robb recently wrote some interesting things about this very matter:


Milky Way 9:38 AM on 19 Sep 2008

This very question is asked by Ronald Coase about 80 years ago. Ronald Coase, trying to answer this very question, did work that earn him the Nobel Prize in Economics. You should congratulate yourself being able to notice this wonderful question. His writings on this topic is highly readable. You should try find them on Google. There are, in my opinions, further and better answers. But his original writing is still good enough.

Sometimes I warm up when I see Free Software programmers learn economics.

Jimm'ney Cricket 11:27 AM on 19 Sep 2008

In my view, specialization is part of competition.
1. On a macro level (e.g., industry), competition has the counter-intuitive effect of being essentially a cooperative effort to optimize how to deliver products/services to specific groups of customers. However, even companies end up specializing. Companies end up specializing to optimize their capabilities with the benefits of delivering those capabilities.
2. On a sub-macro level (e.g., department), competition is an optimization of how to use the scarce internal resources for the purpose of serving the organization. Departments frequently compete through management proposals, meetings and budget reviews to determine who gets the funding.
3. On the micro level (e.g., individual, resource), competition is an optimization of who/what best fulfills the needs of the sub-macro level. For example, last week our HR department reviewed 1200 candidates for a single IT position...which certainly sounds competitive to me!

Ian Bicking 11:29 AM on 19 Sep 2008

Someone was telling me that Motorola kind of has internal services like this. Departments have special Motorola Bucks (probably they don't call them that ;) and they can spend them on some competing internal services (but only internal). Universities, where departments are often quite autonomous, often can purchase internal services with straight cash, or self-provide those services, or get external providers. Internal providers have the benefit of knowing the system, but there's still a full choice. The move towards outsourcing (not necessarily internationally), and the idea of "focusing on core competencies" also alludes to this -- making a company truly one entity, where other services are market-driven.

I think these all have problems because, like a command-economy, the internal decisions in a company are seldom rational and isolated decisions, but are based on a kind of internal cronyism. That said, a lot of modern business planning seems like a way of formalizing things so that the cronyism is harder to hide.

Michael Foord 12:48 PM on 19 Sep 2008


I'd like to see some evidence from John Robb to support his assertion that "Median per capita incomes in the US are the same as they were in 1974 -- there hasn't been any income growth at all".

He has an interesting theory, but in that short article alone he hasn't proven it. :-)

Nicola Larosa 12:56 PM on 19 Sep 2008

@ Michael

Not that hard to find, and I'd say he was right on that point:

Bill Mill 1:06 PM on 19 Sep 2008

jtauber wrote > would it make sense for a VC to invest in two competing companies?

Assume an initial investment X in a company, and that a VC success gains 10-100X dollars and a failure gains 0X dollars. (i.e. consider anything less than a 10X return a failure and ignore those gains for a worst-case scenario). Further assume that chance plays a large part in determining which company wins in a given segment.

From there, once you believe that an un- or under-served market segment has the potential support a company big enough to deliver a return of 10-100X dollars, you trivially want to invest in at least 10 competitors in the field, since you don't know which will be successful and you maximize your chance for a big strike.

But I agree with Jeff that this argument doesn't support the "multiple competing internal entities" model.

William Hertling 1:22 PM on 19 Sep 2008

Gifford Pinchot wrote on exactly this topic in his book The Intelligent Organization. He calls the concept free intraprise, and he's successfully worked with more than one company to implement it.

Pinchot basically works to turn cost centers into profit centers: employees get to form intrapreneurial teams that earn budget dollars from other departments.

Here’s a more detailed example of how free intraprise works:
1) Several employees find each other and realize that they have a complementary set of skills and could form a team: perhaps an R&D team, perhaps an IT team, perhaps a marketing team, or even a customer support team.

2) They decide on the services that they could provide to the rest of the company, and advertise those services internally.

3) When others in the company need a service, they will now have three options. The two old options are that they can get the service from a standard organizational group at no charge (i.e. getting something from your regular R&D group or IT group through a typical POR process), or they can use their budget to outsource the work to a vendor. The new third option is that they can use their budget to in-source the work to a free intraprise team.

4) If they decide on in-sourcing, they are choosing to utilize that internal profit-center team: a group of experienced company professionals, working inside the firewall, without any confidentiality concerns, comfortable with the company culture, with knowledge of the company businesses and strategy, etc.

5) The free intraprise profit-center team takes the budget dollars and pays their own salary and related expenses out of the budget. If they can’t cover their own expenses, then, well, they are out of a job. But as long as they can cover their own expenses, then they are in control of their own destiny.

Pinchot shows that 80% of the work that happens inside a large corporation is providing services to other parts of the company. So this actually has quite a bit of potential. There are numerous benefits, including keeping talented people inside the company, keeping intellectual property inside the company, the reduced friction of working with an in-source free intraprise team rather than an out-sourced company, etc. It can do tremendous things to reduce bureaucracy and enable huge leaps in productivity.

Ali 3:36 PM on 20 Sep 2008

I think James pretty much answered it -- Coase's original question was similar to yours, roughly: if markets are so great, why do we have these "little islands of communism" (aka firms)? Basically, markets beat firms when transaction costs exceed possible benefits from managed effort.

I wonder if your scenario is more plausible with technology. How would you lower the transactional costs? The labor market (consulting, outsourcing) aren't always great substitutes for the IT department because of the transaction costs of discovering services, negotiating contracts, enforcing contracts, ensuring settlments, etc. Otherwise, all scenarios for the Red/Green/Blue teams can already be achieved through consulting and outsourcing.

You could emulate financial markets (NYSE/NASDAQ for internal IT), and have predefined contracts for IT tasks like hosted servers, advertise them, ask, bid, close and settle. But how many IT scenarios can this cover and what are the costs of negotiating and enforcing ad hoc (over the counter?) contracts? And how to develop the market infrastructure for discovery, contract enforcement, trade and settlement? While being cheaper than how the firm currently engages in such efforts?

zoura 8:23 PM on 20 Sep 2008

in most cases, competition means: concentration of resources, distribution of losses.

Zoran Lazarevic 4:17 PM on 22 Sep 2008

The answer is: transaction cost.
You can find the full explanation at always excellent Econtalk: see podcast Munger on the Nature of the Firm

Executive summary: The companies should never hire anyone, because the owner could contract a few managers temporarily who will outsource design, manufacturing, marketing ... to freelancers for as long as they are needed. But the transaction cost of finding a freelancer every time you need to open an email account is too great. Think also about the human capital (knowledge about your specific organization), and availability.

Michael R. Bernstein 8:00 PM on 27 Sep 2008

A related question: Why are most firms organized as fairly rigid demi-meritocracies with limited vertical mobility and an aristocratic class that is shared across many firms?

I am not advocating Co-Ops, they do exist and have their own problems that prevent the form from being more popular, but where are the firms organized like Constitutional Democracies?

Add a comment:

Ignore this:
not displayed and no spam.
Leave this empty:
not searched.
Name and either email or www are required.
Don't put anything here:
Leave this empty:
URLs auto-link and some tags are allowed: <a><b><i><p><br><pre>.