Wednesday, December 07, 2011

Private Sector Incentives

A Northwestern grad student (just like I used to be) blogs on academic papers over at A Fine Theorem, and I was struck by this snippet:
The benefit of capitalism can’t have much to do with profit incentives per se, since (almost) every employee of a modern firm is a not an owner, and hence is incentivized to work hard only by her labor contract. A government agency could conceivably use precisely the same set of contracts and get precisely the same outcome as the private firm (the principle-agent problem is identical in the two cases).

This is wrong. I've worked in small and large firms, and in either case, being part of any initiative that makes money increases your pay and stature rather clearly. Now, the bigger the firm, the more people involved in any initiative, and the more people will exaggerate their involvement with successful projects and downplay their involvement in losers. But the bottom line is the bottom line, and the best way to get your boss to appreciate you more is to help him make more money for the firm. The scope of your command is rarely specified in detail, though increases in scope are the most obvious ways of increasing your power when the 'contract' comes up for renegotiation: work is a repeated game. The government principle has a very different objective function than the private firm principle, the former with a diverse set of stakeholders to satisfy, often with vague and inconsistent objectives, the latter to simply make more profits.

It's true that some people, especially those in over their heads, who are overpaid, dislike productive and smart direct reports because they sense their boss will recognize an attractive replacement. Yet even here the incentive clearly exists for this bad boss to be replaced by his boss, because his boss is not optimizing this position in that case (and watching the bad bosses employees leave when they recognize this is a really good signal that their boss is not a good one). It's also true that salaried employees like cashiers and secretaries may not be proactive about increasing their stature within the firm and simply follow orders, but their bosses recognize this and so are less likely to increase their responsibility.Every ambitious worker tries to make more with less all the time, because they can quantify this, which is the best way to go into an annual performance review. Just because you do not have shares in the firm does not mean you aren't highly incented to make money for shareholders, because the shareholder desire for more profits truly trickles down.

If this kid doesn't appreciate this, he's missing a big part of what makes private organizations work as they do. The nice thing is that it isn't a really fragile result like some of these theoretical models based on the types of agents and some envelope theorem. Common sense tells you the more you appear to be creating profits, the better your future will be in the firm and the industry, and the best way to appear to create profits is to actually do so. Economics, as Tom Sargent notes, is 'organized common sense.' Unfortunately, while you can formalize and thus teach models, you can't formalize and teach common sense, which is why game theory is so ambiguous on these important matters.


Kevin Bryan said...

Hi Eric,

I'm the author of the post you mention. I actually think we don't disagree at all. Both state-owned firms and shareholder-owned firms hire people to perform some task in order to meet some objective (probably profit maximization for private firms, and probably some combination of cost minimization and other social welfare considerations for state-owned firms). Since the private sector workers do not directly own the firm, they benefit from increased profits only to the extent that their contract causes them to (and I mean this in the broadest sense, so reputational contracts and other repeated game concepts are fine).

The point is that a private sector worker can be incentivized by payment schemes that link his effort to his pay, but that such schemes are internal features of a firm's organization, not features of private sector capitalism per se. Likewise, government workers can be incentivized to cost minimize simply by giving those workers exactly the same nexus of contracts they would get in the private sector. Now, if you think there is some reason why private sector firms are *more likely* to give out such incentivizing contracts (public choice, or the type of evolutionary argument I made in my post), then I am totally with you. But then the benefit of the private sector is that various aspects of that sector tend toward contracts that link worker profitability with effort, which is a very different thing indeed from the common belief that workers in the private sector are uniquely and inherently motivated by the private sector.

Eric Falkenstein said...

Kevin: Well nice to meet you (in cyberspace). Say hi to Mark Satterthwaite for me if he's still there, he was always very generous with his time for me and I appreciated it.

You say:

"private sector worker can be incentivized by payment schemes that link his effort to his pay, but that such schemes are internal features of a firm's organization, not features of private sector capitalism per se."

I think an incentive contract linking someone's payment to profits is the essence of capitalism; the distinction between public and private firms.

"Likewise, government workers can be incentivized to cost minimize simply by giving those workers exactly the same nexus of contracts they would get in the private sector."

First, they can but they don't. Why would government workers cost minimize? They usually measure output via their costs, which is why the ratio of admin-to-teachers has risen over time in public schools. Look at all the economists(!) who look at 'G' and simply assume that it is identical to any 'I' spending. A manager of G is on a perpetual political campaign, and while there's some lip service to efficiency, most of it is based on jobs and money spent, and the demographics of its impact.

So, just as Kantorovich showed, it's possible to incentivize socialist collectives to behave just like private firms, but as they have different objectives by definition, the incentive structure for private firms will be different, so they won't behave like private ffirms. This is why when the East German wall fell, their signature auto, the Trabant factories were simply abandoned because they were so crappy.

I think you are getting confused focusing on the fact that govt and private workers are often paid a salary or hourly, because mentioned, these are mini contracts in a longer, repeated interaction, where principles like agents who make more profits.

On bigger issue, I think there's something very rotten within the whole concept of duality, because minimizing costs is not just like maximizing profits, but I don't have the key proof at hand. But the essence is in Hayek's 'Use of Knowledge in Society' paper, that firms owners, and all the myriad profit maximizers within their firms, are not facing an obvious, finite set of cost and revenue functions.

Dan Carroll said...

There are two elements to consider as well. One is the darwinian aspect of free markets - firms that fail to organize properly around profits simply fail, or at best stagnate. In government, the darwinian nature works in reverse -managers that downsize and focus on efficiency encounter political resistence and lose status. The empire building impulse is encouraged in government rather than restrained. This creates contrasting cultures, one of efficiency and one of unlimited mandate.

That said, profit maximization is not the only incentive for private sector employees. Maximizing profit share, job security, status are all key motivations that often conflict with profit maximization.

Anonymous said...

Kevin is a foolish child who has to read, among many things, Hayek.

Kevin Bryan said...

I still think there's some confusion here in what I mean. I agree completely that owners of firms and leaders of governments have totally different objectives. What I disagree with is that *workers* at firms, and *workers* in governments have different objectives that result from firms being profit maximizers.

As Dan notes, workers in firms are not owners and hence do not necessarily have profit maximization as their goal. Surely empire building, job preservation, sloth and many other factors affect private sector workers unless they are incentivized otherwise. And the same factors also affect government workers unless incentivized otherwise. I don't believe the distinction between public and private is that private firms can write contracts linking firm profit to worker pay - the distinction surely is that one sector is government owned and one is not (if you're a Hayekian, this just means that the separation theorem applies to the principal in one and not the other). Given that distinction, you might ask why private sector firms are more likely to use high-powered contracts which link pay to output? I argued in my post that this is generally a result of selection effects that result from market competition.

That statement is radically different from believing that private sector firms somehow have a magic tool to incentivize workers which government does not have. In particular, it makes me wary of monopoly power - and does anyone doubt that monopolies tend to also see the type of sclerotic work force you often see in the public sector?

I think a lot of the confusion here is coming from your giving agency to firms or collectives, rather than individuals. A "firm" does not have incentives at all. Owners do, and managers do, and workers do, and coordinating those is very difficult. Indeed, Mark (who is still here and is still a very nice guy!) has discussed this point at great length in his own theory.

Mercury said...

Kevin - The magic tool of private sector firms is that they can fairly easily fire workers who underperform. Much more rarely (if at all) are public sector workers ever sent packing for poor performance. Likewise, the private sector worker is in a much better position than his public sector counterpart to pick up and go work for the competition if he feels that his efforts aren't being properly rewarded. These may be more implicit than explicit "contracts" but they are important and powerful incentives nonetheless.

For sure, the larger a private firm gets the more workers there are at that firm who behave more like public sector workers (building fiefdoms, garnering perks, becoming detached from the core mission etc.). In fact, once a firm achieves a certain mega-size, they practically become government appendages (see TBTF banks). All the more reason to try and keep the economic climate as small business friendly as possible if you want it to be dynamic and robust. More small businesses also means more principals and less agents too - so you reduce the size of that problem.

Eric Falkenstein said...

Kevin: I don't understand why you think workers at private firms are not incented by profits. As I mentioned, that is the best way to enhance their career, even if they are salaried or hourly employees, and this is the main difference between public and private sector employees.

Patrick R. Sullivan said...

Hi Kevin, you can say hello to another Mark at Northwestern for me (who was a grad student with Eric when he was there). I used to tease my friend that he must have had a lemonade stand or a lawn mowing business when he was a kid, because unlike a lot of academic economists I knew, he could appreciate real world applications of his discipline. He finally admitted that he'd had to work in payroll at his family's box making factory when he was young.

You might profit from discussing your ideas with him, because it's pretty clear you don't have such experience. Not trying to be rude, but you need to realize that there are limitations to what math can capture.

Eric is correct that even employees who aren't explicitly responsible for bottom line results will have that responsibility passed on to them through their everyday interactions with their co-workers and managers. You might say they'll be led to it almost by an invisible hand....

Ken said...

I think the author has too broad a concept of "contract". Actual contracts are quite short, most of the implied terms depend on the parties assessment of each other's motives/habits.

Government has different motives/habit than private sector and so the implied terms will never be the same. You can't just transplant the "contract" from one entity to another.

Unanimous said...

I work for a company that provides "outsourced" services to governments. We replace small sections or departments of governments with more efficient operations. We are typically 20-30% cheaper for the same service, and we do the service significantly better. The way we do the services better is through having better systems (technology and procedures), and we need these to meet contractual service levels. And we are cheaper due a combination of the better systems and paying employees less.

Governments can't put the types of contracts in place within their own organisations that they can with an external company - they only have employment contracts. The higher up part of the government gets more control of their operations by outsourcing with commercial contracts in place than they do by keeping it in house (which often strikes people as counter intuitive). The operation and it's objectives are legally defined and made clear to the people carrying it out, and also made clear to the contract managers within the government. Government contract managers don't usually come down hard on us on the few occsions when we don't meet our SLA's, but we know what we've done, we know that everyone involved knows, and we know we need to fix it.

There is little difference between our front line employees, and government front line employees. There is little difference between our managers and managers with equivalent roles in government. We pay a bit less, and seem to end up with people who don't mind doing a fair day's work for close to a fair day's pay provided they like the people they are working with. People who want more pay end up finding an equivalent government job. I don't think that financial incentive for employees has anything to do with why we can do better than governments.

In a way, we are part of government, but a demonstrably more efficient part, and, we are also a company.

It's my view that there is a lack of direction and clarity within governments over what they are trying to do. Some of that may be due to size, but even small local governments show the same behaviour. Most companies seem to have a fairly well defined market and a profit objective. Every government is trying to provide lots of services while balancing popularity with un-popularity across the population. The less well defined role I think is a big part of the difference.

If financial incentives were a big deal for employees, then employee owned companies would have out competed non-employee owned companies by now. The fact that they haven't is telling. They have been legally possible for just as long as non-employee owned companies, and enough have been tried and survied for long enough to show that they can function, but they have never dominated in any economy.

Kevin Bryan said...

Eric and Patrick,

I do think private sector workers can be incentivized by profits. We don't disagree about that. But the incentive is indirect, not direct. If the private sector workers are incentivized in a way linked to profits, it is because the owner uses contracts (in the broad sense, as Ken notes) that make it so. But this is a choice by the owner, just as it is a choice by government to focus on maximizing things other than profits for state-owned firms. Why those different types of contracts/relationships tend to appear is an interesting question (Unanimous notes that government may be limited in the types of incentive schemes they can use, I focus on selection pressures in the market, etc.). The distinction is much more subtle than just that "private sector workers care about profits, and public sector workers don't, and nothing can change that."

But this point matters! If workers at firms care about profits *indirectly*, then such incentives may differ in different industrial structures; monopoly versus competitive markets is the distinction I wrote about, but there are others.

Unanimous said...

Eric, you said:

"I think an incentive contract linking someone's payment to profits is the essence of capitalism; the distinction between public and private firms."

But very few employment contracts are like this. If what you say was the case then most employment contracts would pay people a proportion of the company's revenue. You can find plenty of employement contracts with some incentive payments, and some extra-contractual incentives as you've identified above, but it's very rare for a company to fully pass on its incentives to its employees.

In fact it's very interesting the extent to which companies don't pass these incentives on. The range of possibilities is huge - you could even in theory exagerate the company's incentives in an employment contract. But if you look at what actually happens, it's almost as if one of the reasons companies exist is because they buffer employees from market incentives and risks to a large extent.

I think Kevin has a point, things are much more subtle.

Eric Falkenstein said...

Kevin: I don't think it's so subtle. The owner wants profits, his direct report wants to please the owner (to get a bonus, not be fired, to get a higher salary next year), all the way down. In government, the objective function basically excludes profits, and this incentive goes all the way down. I guess we are talking past each other.

Unan: the renegotiation of the salaried worker implicitly makes the employee incented by profits.

Michael said...

My experience is that private sectors employees are motivated to avoid job loss, not to make profits for the owner per se. However, they do know, often first hand, that no profits means no job.

Mercury said...

That was one of my points Michael. Keeping oneself on the left side of *have job/don't have job* is often incentive enough (especially now)to please the profit-motivated boss.

There's been some talk here about private sector principals sharing business profits with workers further down the ladder...but nothing about sharing the risks.

Some workers want more risk/reward, some want more security. Those workers for whom security is really a priority tend to go work for the government.

Dave Pinsen said...


"Kevin: I don't understand why you think workers at private firms are not incented by profits."

Because the majority probably aren't, since they don't share in those profits in any material way. I think this even applies to companies that have profit sharing plans. Does the average call center or warehouse worker at a mega corp do his job any differently knowing that he might get an extra 2% match in his retirement plan if the company meets a profit target? Also, sometimes what makes the company more profitable (merging with another one and eliminating redundancies) increases the chance they'll get laid off.

The bigger difference, in practice, between public and private sector workers is the one mentioned by a previous commenter, that public sector workers are harder to fire. It's not the profit incentive keeping the average (non-union) private sector worker on his toes -- it's the prospect of getting summarily fired.

There are some government employees who have a reputation for competence and effectiveness, despite profits not being part of the equation. FEMA may have dropped the ball during Katrina, but the Coastguard and National Guard operated effectively. That would seem to support Kevin's argument: the salient issue for workers isn't profits at the organizational level but incentives at their level. The military and the Coastguard have a broader range of incentives and disincentives to apply than a civilian government organization does. They can summarily punish their workers, for example, in ways that civilian government organizations can't.

Leo said...

"I've worked in small and large firms, and in either case, being part of any initiative that makes money increases your pay and stature rather clearly."

What makes you think that the handful of experiences you've made in the bonus-laden world of quantitative finance (according to your resume) is in any way representative or even statistically significant?

Your blog entries are full of these kind of fallacies.

Eric Falkenstein said...

Leo: If you think most bankers are Investment bankers with large bonuses, you would be wrong. Goldman Sachs employees aren't typical.

Leo said...

Don't be silly Eric, I hold a position in quantitative analysis very similar to some you describe in your resume. We enjoy salaries and bonuses that are in the upper few percentiles. Generalizing our experience is dummheit and/or constitutes intellectual fraud.

Anonymous said...

For those interested, TexMetals has a great promotion going on sealed boxes of 200 Grizzlies in Thermatron packaging. Spot is $7.95, which is actually really good for the grizzlies (APMEX is $10 over). Only 2400 left. If I had any cash I would buy.