NOTE: The Growth Economics Blog has moved sites. Click here to find this post at the new site.
You have to be careful in confusing “free markets” with “perfect competition”. By “free markets”, I think we mean free entry for new firms and/or products into the market. We don’t want restrictions on innovators from bringing their ideas to the market. We typically *assume* that free entry exists in economic models, but one thing that holds back development may be the absence of this free entry (think red tape and bad institutions).
But we don’t want “perfect competition” even if we do want “free markets”. One of the counter-intuitive things that comes up in growth courses is that perfect competition is not conducive to rapid growth. The story here involves a few steps
- Growth is ultimately driven by innovation
- People will innovate if they have incentives to innovate
- The incentive to innovate comes from economic profits
- Profits only exist when the innovator or firm has some market power
Innovators and/or firms need to charge a price greater than marginal cost to earn profits, otherwise there will be no incentive to innovate, and ultimately no growth. If you allow competitors to copy innovations they will drive the price down to marginal cost, eliminating profits and incentives for innovation. We want free entry of new firms with market power, but not free entry of imitators who produce perfect competition.
But perfect competition does maximize the combined consumer and producer surplus from a given product. So there is a tension here. Perfect competition maximizes the output of *existing* products, but minimizes the output from *potential* products. Think of it this way, if we decided that we had all the types of goods and services that we could ever want, then we’d want to enforce perfect competition. We would nullify every patent, and let competition take over to maximize the output of those existing goods and services. Nullifying patents (or any other kind of intellectual property) would crush the incentives to innovate, and we’d never get any new products.
This means that it is not obvious what the right policy is for intellectual property rights and/or competition in general. It depends on your long-run perspective. You can trade off long-run growth for a higher level of current output by canceling intellectual property rights. Or you can trade off current output for a higher long-run growth rate by enforcing property rights strictly, and probably instituting even stronger ones.
There is no *right* answer here, because it depends on your time preferences. But extreme answers are probably unlikely to be optimal for anyone. Strict perfect competition – allowing imitators to ensure P=MC – isn’t good because it prevents us from getting new products. Super strong market power – limiting each good to being produced by a perpetual monopolist, say – would shrink the availability of every existing product, even if it makes the incentive to innovate huge.
The Growth Economics Blog 
Be careful here – quasirents do not require limiting free entry! For lots of types of innovation, the fixed cost to innovate must be paid by every firm including imitators (or alternatively, learning by doing gives MC advantages to early movers). Agree that this means that firms will wind up some market power, but do not agree that growth requires artificially generating that market power (in almost all cases).
I think the point still holds. If there are fixed costs, then sure, this limits entry. This creates “rents” in the sense that entering firms need to get paid back for eating the fixed costs. But those are not rents to innovation, those are rents for eating fixed costs. You need some rents to innovation to incent innovation. Why invent a new product if all you can do is earn enough to cover fixed costs?
That being said, I don’t think we need to artificially create market power, unless by artificial you mean patents and other IP. But we shouldn’t be granting perpetual monopolies and such to anyone, that’s for sure.
Pingback: Links for 11-16 -14 | The Penn Ave Post
Well said. But I would add an important nuance here. The drive to escape perfect competition can be an important motivation/incentive to innovate. The product cycle consists of two broad stages: (1) product innovation, where new products are introduced, and the typical form of competition is in differentiation; in this stage, firms can earn a positive profit, and (2) process innovation, which is when markets are mature, and competition is in cost and the outcome most closely resembles the Bertrand model. The desire to escape Bertrand is what drives innovation.
So the issue isn’t one of “perfect competition bad for innovation,” as much as it is the rate at which market structure transforms from perfect to imperfect competition influences both the pace of innovation and the rate of diffusion of that innovation towards an efficient outcome.
Right on, Seth. The post is very 0/1 about competition, but as you say there is a little subtlety here. Firms always want to keep market power, and avoid competition. The policies and rules we make need to allow for some market power to incentivize innovation, but not so much that firms can continually escape the Bertrand competition. Point is that there is not a definitive answer on the right level of market power.
Completely ignores the fact that monopolies destroy consumers savings leaving no incentive to enter the market because there aren’t any consumers. Nominal growth is achieved by spending, real growth is achieved by breaking monopolies.
So this is exactly the trade-off we’re talking about. If you cancel every patent, copyright, etc.. then yes, you would get lots of cheap versions of everything. But you’d cut off the incentive for future innovation. So what’s the right line? It’s not the case that companies having market power is universally and always bad.
to widen the market and narrow the competition is always the traders goal, there’s no need to provide bounties. The traders will diversify unless they have reason not to, like being granted the privilege of taxing the rest of society. All it’s needed is demand which we don’t have because of all the anti-social supply side policy. Also, instead of expecting the real world to speed up to match the financial world, which is folly, we should be slowing the financial world to match the real world.
I don’t think the point that cancelling patents, copyrights will discourage innovation. Perfect competition leads to zero economic profits, but not zero accounting profits. Firms still earn a required rate of return on their products. What they do with those accounting profits is entirely based on the firm’s investment decision – i.e. reinvest in current business, declare dividends or invest in further innovation. Patents and copyrights are like the cream on top, which may or may not be necessary to promote innovation.
So what if accounting profits are non-zero? There is a reason we distinguish between accounting profits and economic profits. Acct. profits include economic profits as well as the implicit payments to capital. If you drop the economic profits to zero, then acct. profits will be sufficient to pay for renting capital, but no more. People invest in innovative firms to earn rates of return *higher* than they could get just buying an index fund, say. So why would anyone invest in a new innovative firm if they could earn the same return in an index fund with far less risk?
Hello Dietz, I really enjoy your blog. I have some questions about this article.
I just read an article with Ha-Joon Chang and he suggest that tariffs for developing countries helps them to create in the long run better economies / growth. His example is that Japan at the beginning had a very weak car industry, who produced only 70.000 while GM at the same time 8.5 Million and only through this tariffs the Japanese car market was able to compete in future (now). He argues that otherwise Japan’s car industries would be eradicated in an open market. Does this mean that competition and growth not only rules within a country only but as well on the international level of the market? Which lays another layer of trade-off in my opinion. Where do you see Schumpeters-Growth-Model at the open market level?
Furthermore, innovation from the public sector like NASA or Universities with their fundamental research don’t they innovate for the researches sake only? I haven’t heard any Nobel laureates who hold a patent on their research e.g. the recent winner in physics who invented the blue LEDs while competing and succeeding against the private sector. I guess in general the difference is the commercial use for some of the innovation?
Best Grüße
Leath
This is a good example of how neoclassical ideology gets economics wrong. It makes simple assumptions and extrapolates the wrong conclusions.
For one, profit is far from the strongest motivator to innovate. The workers who do the actual innovating (scientists, researchers, engineers, etc.) have different drives. According to Dan Pink, the research says people are motivated by autonomy, mastery and purpose.
Two, a lot of innovation stems from public investments in basic research.
Three, open-source software has made a significant contribution to innovation without a profit motive.
Four, there is no direct relationship between stronger property rights and stronger long-term economic growth. Certainly patents provide an incentive to innovate. But long time-period patents can impede the pace of innovation (innovation is built on top of innovation.) Frivolous software patents get in the way of software innovation.
Five, patents only protect big corporations with an army of lawyers. Small players are not only not helped by patent laws, they can be preyed upon by patent holding companies. Blackberry was forced to settle a lawsuit from NTP for $600-million.
Six, as Joseph Stiglitz points out, monopolies like Microsoft can use their market power to usurp innovation: “It did not develop the first widely used word processor, the first spreadsheet, the first browser, the first media player, or the first dominant search engine. Innovation lay elsewhere. This is consistent with theory and historical evidence: monopolists are not good innovators.”
Seven, competition also leads to innovation. As the saying goes, necessity is the mother of invention. Compare broadband internet between North America — which is controlled by oligopolies — to Europe where basic telecommunications infrastructure is regulated by government leading to more competition. According to the Harvard study Next Generation Connectivity, the European model leads to greater innovation (as well as lower prices.)
In short, some degree of patent rent is good for innovation. But more is not necessarily better and rent-seeking corporations can abuse the process.
So you have to actually read the whole post. Neither perfect competition nor monopolies are good for innovation. You have to get the mix right. The answer to patent trolls isn’t to cancel all patents, it is to get the patent law right. You want some IP protections, but not absolute protections. There is a balance between perfect competition and absolute monopolies. That’s what Stiglitz was talking about – MS was essentially a huge monopolist and drove down innovation. But the answer to that is not to cancel every single one of MS patents or copyrights and let people mimic them for free. That would destroy the incentives of the next MS to even get started.
This is a depressing post which takes models as if they were good descriptions of the world – rather than descriptions of an imaginary world built to help us clarify our thinking.
Anyway, if you’re interested, here’s a post by me suggesting that something close to perfect competition on the internet generates huge innovation opportunities – but it’s not in a worked out formal model.
Not sure what you mean by depressing, and there really isn’t a model involved here at all. It’s a simple story: if people can copy your product and drive your profits down to zero, then why bother inventing the product at all? And if you want some backing for that, there are papers out there showing that innovation (R&D spending and new products) are lowest when an industry is either super-competitive or super-concentrated. Perfect competition, empirically, does not lead to faster innovation.
It’s a model of something else – transactions costs – though discursively intimated – not formally spelled out – not sure what the point of spelling it out would be.
It’s pointing to something which may be more important in innovation than the this post’s focus on the margins necessary for investing in R&D. A lot of innovation doesn’t come from R&D. My ‘story/model’ focuses on transactions costs involved in making connections and doing business. If innovation is a quality of systems rather than firms, then the focus of this post may not be very germane to the issue.
Still that wasn’t the burden of my main point here which is that the post shoehorns the world into something which it would need to be to make the models relevant. There are a few areas in which this is the case – like pharmaceuticals. But huge swathes of innovation and growth are not sensibly conceptualised as involving substantial R&D which requires monopoly rents to be funded and is then commercialised. Of course you can use the model of perfect competition to prove – positively prove – that it’s the case that you need monopoly rent to spend the surplus on R&D and then think that anyone who disagrees with you just doesn’t understand the model. I do understand the model, but I think it doesn’t help us think about the world of innovation except for quite specific cases – such as pharmaceuticals as described above.
I understand your point better now. You’re right, there are sectors/firms/industries/technologies where the patent/IP model of innovation isn’t as applicable. And in that case the relationship of the level of competition to innovation will be “flatter”. But I’ll stick with the main point, which is that perfect competition – in the sense of allowing other firms to copy you costlessly – would lower the incentives to innovate in any dimension. Lower the incentives to zero? Probably not. So growth wouldn’t go to zero. But it would likely be lower.
Pingback: G7+ zamiast G20
Okay, but perfect competition does not even exist. Its like discussing the merits and drawbacks of teleportation in a transportation engineering class.
Which reminds me of an post by Manski on the fiction of deadweight loss (a corollary of the fiction of perfect competition).
http://www.voxeu.org/article/removing-deadweight-loss-economic-discourse-income-taxation-and-public-spending
Absolutely – but take the post as saying that even though PC doesn’t exist, we don’t *want* it to exist.
Hello Dietz, I really enjoy your blog. I have some questions about this article.
I just read an article with Ha-Joon Chang and he suggests that tariffs for developing countries helps them to create better economies / growth in the long run. His example is that Japan at the beginning had a very weak car industry, which produced only 70.000 units while GM at the same time 8.5 Million cars and it was only through tariffs that the Japanese car market was able to compete in future (now). He argues that otherwise Japan’s car industries would be eradicated in an open market. Does this mean that competition and growth not only rules within a country only but as well on the international level of the market? Which lays another layer of trade-off in my opinion. Where do you see Schumpeters-Growth-Model at the open market level? Obviously economist always talk that Japan specialises on their products and US on theirs but this blocks the idea of long run future competition of the same products! Answers would be really appreciated.
Furthermore, innovation from the public sector like NASA or Universities with their fundamental research don’t they innovate for the researches sake only? I haven’t heard any Nobel laureates who hold a patent on their research e.g. the recent winner in physics who invented the blue LEDs while competing and succeeding against the private sector. I guess in general the difference is the commercial use for some of the innovation?
Best Grüße
Leath
Leath – thanks for reading. On tariffs and trade, there are several people who propose that tariffs are necessary or optimal to get industrialization going. Too far in that direct and you get “import substitution”, which doesn’t seem to work. But being export-oriented – focusing on producing goods that you sell to other countries – and then moving up the value chain as you get experience, does seem to work. Think Japan, Korea, China.
And absolutely, a lot of research gets done publicly, but then brought to market by private companies. Think of there being some given flow of basic research that could be publicly funded, but then the innovation decision based on profits is companies coming up with applications of that research. But yes, lots of people invent stuff for no monetary gain. So maybe if we had perfect competition innovation wouldn’t go to 0, but it would go down.