November 30, 2013

Production as Privilege: Markets and Non-Markets

I realized when starting a sort-of summary piece on all the Production as Privilege series of posts that I had skipped straight from conceptual tool 7 to conceptual tool 9. That was enough of an excuse to go back and finish up with one more idea I had been wanting to write about. Here it is.

Conceptual Tool #8: Market and Non-Market Domains

The market is not everywhere. It may affect in some way nearly every domain of our lives, but there are social domains where other forces exert far more influence, and there are also social domains where we have set up deliberate barriers to keep the market out. Some of these places are surprising; one of the most surprising to me, when I first encountered it conceived in such a fashion, was the firm.

It may seem odd to think of firms (corporations in any form) as anti-market structures, but in fact that is exactly what they are. Firms are set up to keep market forces at bay, as decades of "theory of the firm" literature started by the late Ronald Coase has shown us. The "interior" of a firm is not a market but typically some type of hierarchical bureaucracy.

For example, the first person on the assembly line does not sell their product to the second person. They simply pass it along; the product is owned by the firm and the workers have agreed in advance to sell their labor. Similarly, most office workers do not have to buy own their computers--they are simply assigned them. These non-market workplace setups happen for a number of reasons: it would be very inefficient if each worker had to agree on a price before they passed along the product; or businesses may be able to get a better deal on a bulk computer purchase.

Likewise the allocation of work within a firm is very different from external labor markets. Workers may compete with each other for promotions but not to get paid for the work they are assigned. Workers perform various tasks for the firm without negotiating each price.

This all serves to remind us that capitalism has always had a more complex relationship to markets than we often think. The economics discipline tends to idealize markets and assume they are ideal in any situation, but this view ignores the myriad of daily interactions and relationships that have nothing to do with markets, and are governed instead by hierarchy or other social arrangements.

Gradually, thanks to a large extent to capitalistic profit incentives, the market has come to intermediate a good deal more than the basic exchange of commodities--not just our daily labor but our nearly everything we do, from cleaning our house to keeping us in shape. It is not a one-way process, however. Markets often recede naturally when they are inconvenient: for example, my house is currently managed by a rental company, so I do not have to contact the repairman or pay for repairs when my sink stops working.

Curiously, pre-capitalist relationships of production may have been more directly tied to the market than current ones are. Braverman argues that before capitalism, the division of labor was entirely confined to divisions between products, meaning that one person was responsible for the entirety of a product that was sold on the market. Their labor was therefore expressed in the market through the market value of the commodity they produced, instead of through contractual agreement with owners of commodity-producing processes. Over time, however, economies of scale and other forces pushed workers into salaried positions in hierarchical organizations.

What does this mean for our system of production and the way it distributes wealth? By highlighting the non-market nature of the firm, it allows us to question the assumption that workers in a company earn their marginal value. When economists model labor markets, they tend to assume workers are paid the same amount of money that they help a company earn. But outside of a simple econ101 manufacturing scenario, it is much harder than you might think to figure out what the value of a worker is for a firm, because the firm is not a market--the firm exists between the labor markets and the product markets. While the cost and benefit analysis firms employ to decide whether to hire or fire workers is certainly complex, it is mediated by hierarchy that may have little to do with any "value" in an economic sense.

For example, imagine two low-level employees join an organization with the same qualifications. One of the employees is related to the boss, so he is asked to help out with some administrative duties after hours and is eventually promoted to manager and makes significantly more money than the other employee. Now, a manager may indeed provide more value to the organization than a low-level employee, because the manager is more important for the overall running of the company. From this angle it makes sense that the employee with the connections is now making more money than the other employee. However, it is difficult to argue that this outcome is in fact due to market forces. Instead, it has been shaped by personal relationships, hierarchy, and bureaucratic organization.

Ultimately what is important is that we understand ways that markets work and fail to work. Too often they are assumed to be an ideal form of relationship when in fact they are ill-suited for a particular situation. By starting to note all of the places where markets have and have not taken hold, we can get a better idea of how to design policies that use--or do not use--markets appropriately.

July 14, 2013

Public and Private and Other Simplifications

In a society, to a certain extent, our institutions and organizations are created out of thin air: out of our beliefs about how our society works. This means the way we understand things can change their existence, and hopefully better understanding of our institutions and organizations can help them work better. I think one of our fundamental dichotomies today misses the point.

We constantly juxtapose public and private actors: Socialism and capitalism. Government and industry. Bureaucracy and markets. Citizens and customers. They form a dichotomy that we see as inherently different, if not opposed.

But is the difference more than skin deep? Both public agencies and private corporations are organizations--entities coordinating human action. Both are established as a means to an end, to serve a purpose. We have government, and we have private corporations, because they do useful things.* They are tools. Useful technologies.

This broad similarity gets lost in the convolution of various arguments and disciplines. In legal terms, public agencies are established to fulfil a number of different purposes--whatever the laws say that established them, or the government wants, or their board members agree upon. Private corporations are simply a very specific type of organization, with established ownerhip rules and purpose (make profit and minimize risk).

But these legal definitions are just ways of facilitating and motivating organization, they say nothing about what role an organization plays in the actual lives of actual people. Private corporations can provide drinking water and voting booths and go to war with your enemies. Public agencies can make money by providing value in the marketplace.

Now before you get all flustered, I'm not saying they are equally good at any of these things. But that is exactly the point: we have to have reasons for saying that public organizations should do some things and private ones should do others. And we tend to skip those. Frequently.

There are a lot of valid reasons, and they are interesting to think about. But we get so wrapped up in either political affiliations or theoretical crutches (markets are always good! corporations are always bad!) that we forget to think about what the actual differences are.

In terms of structure, hierarchy, and functionality, public and private organizations can be basically the same. They may attract different people, they may have different management styles, but these are variations on types, not fundamental differences.

Obviously the management of organizations is one key locus of difference, and one way to think about this is through the idea of accountability. Public organizations in democracies are ultimately accountable to voters while private organizations are accountable to shareholders (which are often synonymous with "the market"). However, the chains of accountability mean that the buck may in fact stop elsewhere, or simply get passed around ad infinitum.

Accountability is why we give more legal power to public organizations--we would rather have legislatures or judiciaries that we voted for than ones we paid for. Not that there is always a difference, but there most certainly is some sometimes.

Thinking this way puts an interesting twist on common debates about the merits of public versus private organizations, which are typically boiled down to things like efficiency versus greed. For example, the extent of public versus private organizations in a country can be understood simply as whether we want our organizations to be democratically accountable or accountable to the market. Or we can think of the issue in terms of to what degree we want our organizations to act with the force of law.

If we want to design an efficacious society--one that helps us get rich, reduce poverty, fulfill our less materialistic dreams, whatever else have you--we need to understand which organizations are needed, and we need to make sure they fulfill their intended role. If we get too stuck in simplistic ideas of which kinds of organizations should do what (what the state can or cannot organize; what the market should or should not allocate), without delving into the "why", we will not achieve our potential.


*You can dispute individual cases, of course, and you can dispute entire typological existences with arguments like path dependency or new institutionalism or simply power , but I think in the end we have these institutions because they are able to convince us (or at least a sufficient subset of us) that they serve a useful purpose.

June 21, 2013

Production as Privilege: Mercantilism

The Production as Privilege series of posts here has been examining the ways that production connects to the distribution of wealth, through a series of "conceptual tools". Some of the tools are in the basic econ101 toolkit, others are a bit less conventional. Still others, like mercantilism, have been at the core of economic debate for centuries.

Conceptual Tool #12: Mercantilism

For hundreds of years before Adam Smith bestowed his great wisdom upon the land, Mercantilism was the order of the day. In a nutshell, mercantilism was a national policy that believed selling products abroad and running a positive balance of trade was the route to national wealth. If they bought more from you than you bought from them, you got rich and they got poor, because they gave you all their gold. (Gold was better than wine or cheese because it was more fungible and stored value better. Also it glittered.)

However, the mercantilist logic holds only so far as there is a fixed amount of goods and money, and fixed exchange rate prices. Wikipedia has a good summary of the flaws: more modern logic emphasizes the fact that you can consume what you produce (or more accurately, what you get in exchange for what you produce or convince people to lend to you for what you might produce in the future) and that exchange rates will equalize so as to ensure that one country cannot simply take all of another country's money or stuff. This logic is not impeccable either, but it does do a better job of according with the current reality.

What traditional mercantilism gets slightly more correct is the importance of competitiveness, and if you look at the list of typical mercantilist policies at the top of the Wikipedia page, many of them look uncannily up-to-date: export subsidies, promoting manufacturing with research or direct subsidies, limiting wages, maximizing the use of domestic resources, restricting domestic consumption with non-tariff barriers to trade.

Why are we (and moreso other countries like China or Germany) following the recommendations of a "discredited" economic ideology? And what does this have to do with distribution of wealth between countries or within countries?


Partly it is a matter of confusing definition, as we might expect from a term that has been around for a few centuries. It can be helpful to think of mercantilism as not simply as competition for export share, but more generally as deliberate state intervention to bolster a nation's economy. Dani Rodrik contrasts mercantilism not with unrestricted trade (the simple traditional dichotomy) but with a broader economic liberalism--he frames mercantilism as a question of overall government tinvolvement in and direction of the economy, rather than simply applying tariffs with the intent of getting more gold.

From now on, for the sake of clarity, I will refer to "get all the gold", zero-sum mercantilism as "balance of payments mercantilism", and mercantilism grounded in any state action as "state action mercantilism." Balance of payments mercantilism is typically a goal or at least an effect of state action mercantilism, but they are theoretically distinct because state action mercantilism is not inherently about improving trade balances, it is about improving domestic production. State action mercantilism is also roughly synonymous to "state capitalism".

Rodrik also makes the point that capitalism (in the sense of having capital directed by investors) is still feasible under state action mercantilism--we don't need to wholly socialize investment for mercantilism to exist. State action mercantilism can thrive with private capital, as government can still regulate and nudge that private capital in deliberate ways.

Rodrik's definition of state action mercantilism is a bit broad but provides a useful perspective because it helps reveal the assumptions at the heart of mercantilist and free trade/liberal economic ideologies. In Rodrik's view, state action mercantilism assumes that markets can be deliberately improved while liberalism assumes that government is best establishing well-functioning markets and then getting out of the way. Really, this argument is about where to draw the line about government action in the economic sphere and what kind of government action is necessary and desirable.

Any argument about the extent of government action should be about the efficacy of that action, in both the short and the long term. The major debates about state action mercantilism in the past decades have been about the efficacy of the government encouraging certain industries: that is, is government better at anticipating profitable investment than the market?

Obviously the market has a lot of advantages: thousands of highly trained analysts and people with great incentives to make sure their money gets as high a return as possible. But the government has potential advantages of its own: size, time frame, social accountability, and the ability to shape the rules of the game.

Size: The size of the government means that it can direct investment on a scale that private investors have a difficult time doing. And in investment, scale matters. A company may not be profitable without local supplier firms, or industries may not be profitable without whole clusters of complementary industries. Scale is important, and it has gotten progressively more important in the last 200 years.

Time horizon: While some investors are focused on longer-term value, the liquidity of most investment means that many more investors are focused on shorter-term value. Governments can have a slightly longer time horizon. Although elected officials can serve for as little as two years, that is still far longer than quarterly returns. Moreover, presidents can serve for up to 8 and senators for many more than that. Civil servants, as well, may work their entire careers in government and thus have incentives to promote longer-term economic success. Government can be extremely short-sighted as well, of course. The point is that the right combination of incentives is possible.

Social accountability: While individual investors are only responsible to themselves and their clients, governments can in theory be accountable to their citizens. This means that governments have to worry about how the whole economy works as a system, instead of just individual parts.

Rule-changing ability: Finally, governments have the ability to change the rules of the economy in ways that few individual actors can. By creating subsidies, by taxing, by regulating certain procedures, or through a million other methods, the government shapes what is profitable. Perhaps the most fundamental way that the government can change the rules is by changing the price of labor by changing the share of profit that workers are willing or able to receive. Changes in the price of labor can have fundamentally powerful effects on competitiveness.

So is the question of mercantilism simply a question of when these government advantages win out against highly-informed-but-far-from-ideal private direction of capital? This is one way to see it. The argument has been made from a number of different perspectives over the years, however. One of the most interesting positions was from Frederick List.


Friedrich List is a lesser-known political economist from 19th-century Germany who also lived and worked in Pennsylvania. He was influenced, oddly enough, in part by the economic ideas of Alexander Hamilton. Although Hamilton is not primarily remembered for his contributions to economic policy, his 1791 Report on Manufactures as Secretary of the Treasury helped set the course of the US economy over the next two centuries. Report on Manufactures essentially laid out the state action mercantilist strategy the USA followed more or less until the 1970s.

List made many of the same mercantilist arguments as Hamilton in his treatise, National System of Political Economy. Like Rodrik, he contrasts mercantilism with liberalism and broader ideas of free trade. But he does it in an interesting way, arguing that most classical (and now neoclassical) economic theory suffers from fallacies of composition. That is, economic theory failed to recognize that a whole could be something other than the sum of its parts. Arguing from historical examples, List saw that nation-states had without fail been critically powerful players in national economic development.

List further argues that most economic progress has been made through deliberate state policy to promote not simply a nation's economic strength but more specifically a strong, productive manufacturing base. This requires an understanding of both specific national context and also an ability to understand the nation as a system and affect all the moving parts. Those moving parts are not always well aligned:
Nor does the individual merely by understanding his own interests best, and by striving to further them, if left to his own devices, always further the interests of the community. (134)
How you define sound economic policy and correctly aligned incentives, in other words, may vary depending on whether you understand that society is made up of groups. This is because those groups may establish and change collective behavior. One common way this is done is through laws. List argues that, as we already have a state-facilitated market system based on laws, further state intervention can be easily rationalized on utilitarian grounds:
In a thousand cases the power of the state is compelled to impose restrictions on private industry. [...And the state has no right to do so, as long as the actions of private industry] remain harmless and useful; that which, however, is harmless and useful in itself, in general commerce with the world, can become dangerous and injurious in national internal commerce, and vice versa.
The group determination of laws is an example of the rule-changing ability referenced above. However, the point about individual versus collective incentives is different. Countries, because of their aggregated scale, have actual different incentives than individuals and, if properly safeguarded against, these incentives can strongly enhance collective welfare.

As an example, in recent post I made the point that competitors' own actions can shape the availability of what they are competing for. That is, competitors may be competing for pieces of the pie, but the pie may expand or contract based on the actions of the competitors. This is because the fungibility of money converts many different types of competition into competition for value, and we produce value in order to exchange it for other value. This is not a trivial point, it is the crux of Adam Smith's (and all modern free trade supporters') argument against balance of payments mercantilism (see here) and for the productive power of self-interest.

What is oddly missing from most discussions of the "mercantilist fallacy", however, is the idea of incentives. The size of the pie depends on the incentives of the producers--balance of payments mercantilism was initially successful in England and the Netherlands because it incentivized the state to commission and support corporations with incentives to be competitive. But economists eventually ditched balance of payments mercantilism in part because these ideas placed an arbitrary limit on the size of the pie--they failed to recognize that properly incentivized production is bounded only by technology, resources, and organization. Competition is not for resources some set quantity of gold but for the ability to accrue rents from favorable transactions and spend them on plentiful goods.

Let's return to List for a moment. List argues strongly against free trade between nations because, he argues, free trade can destroy the productive capacity of a country. In theory this make sense: if you have moderately good technology and can produce cars for $1000, but somebody else has really good technology and can produce cars for $500, then under free trade only the country with really good technology is going to produce cars. This is fine if we are talking about wine, and you can just switch to producing beer, and sell beer in exchange for wine. But if we are talking about complex industries with dense networks of suppliers and the coordination of specialized technical knowledge (cars need steel and chemicals and computers and hundreds of different parts, etc...), then free trade looks far more dangerous.

And in reality, countries that have successfully grown their economies have mostly used state action mercantilist policies. To be sure, it is not always successful--the general failure of export-led (and state-directed) industrialization was a major catalyst for the neoliberal revolution (Washington Consensus) in development policy. But the examples of successful development mostly involve deliberate state action including picking industries and nurturing them to competitiveness.

More to the point, successful economic policy has involved getting the incentives right. Successful development stories like South Korea show us that state action mercantilism can be successful if the increase the size of the pie domestically--aka the value that is being created in the country over the long term.


What does all this have to do with the distribution of wealth and income? Under fully liberalized trade with flexible exchange rates, the income distribution between two countries should depend wholly on each country's productivity. As we saw, however, productivity depends on having the right incentives for producers, and it can be understandably hard to achieve such a system if the incentives for producers are entirely set by foreign countries--either intentionally through foreign mercantilism or unintentionally through the "natural" workings of the market (this is another main point of List's). Particularly in a world with widely varying levels of productivity between states, state action mercantilism is thus an important tool states can use to "catch up" and even the distribution of wealth.

State action mercantilism, at its core, is a deliberate reshaping of the market that diverts and redirects capital or consumption flows. Mercantilist reshaping can change the current distribution of wealth, but more importantly it can preserve or increase a country's long term productive capacity by stimulating beneficial competition.

But as liberalism argues, reshaping markets not always a good thing. Using state power in markets is dangerous because when you manipulate the behavior of buyers and sellers you lose track of what an un-manipulated market would prefer, and you can inadvertently crush suppliers and demanders that might otherwise happily connect. Worse, deliberate errors (corruption, cronyism, rentierism) are all too easy to come by as state power is turned toward private interests. State power establishes markets but it can also undermine them if the right incentives are not kept in view.

In effect, reshaping markets is what private interests try to do anyway, whether it is through legal means (inventing a new product, changing consumer demand through advertising, or just bringing down prices by selling more of a product) or illegal ones as mentioned in the previous paragraph. Reshaping markets--changing the options that buyers and sellers have, changing their preferences, changing the ways that buyers and sellers interact--is how businesses make profits. States need to be able to establish firm limits on how firms are able to reshape markets and keep the playing field fair if markets are to have any kind of equitable distribution.

But even a "fair", well-functioning market may not end up with a suitable distribution of wealth (if businesses have increasing returns to scale, for example). The state's advantages--size, time horizon, social accountability, and rule-changing ability--put it in a good position to address these further equity concerns. We already do things like provide public education, which is an enormous force for equity. With a broader understanding of how the state creates markets we can hopefully find new ways to make markets fairer--and recognize when markets are simply not the best solution. We must keep List's dictum in mind, that private interests are not always the interests of the group.

May 10, 2013

Competition in Context: Workers

In a recent post post I outlined five points about competition that tend to go overlooked in economic discourse:
  • The amount of things we are competing for changes over time.
  • The future amounts of things we are competing for are uncertain.
  • The things we are competing for may be created by the competitors themselves.
  • The amount of things we are competing for may be artificially constrained (this is basically a roundabout way of describing the way we often create markets using public or non-rivalrous goods).
  • The current allocation of things we are competing for may have important effects on the ability of competitors to compete for more of them (this is basically a roundabout way of describing things like economies of scale, network effects, path dependency, and first-mover advantage).
I want to take these ideas and look at what they can mean for competition situated in various contexts. What does it mean for a country to be competitive as compared with a competitive business or a competitive worker? I don't want to give a full account of how competition in these areas works, rather to think about why the points I've made above are important.

In all three of these competitive arenas, competitiveness comes from offering more value than competitors are offering. Typically 'offering value' means selling similar (aka substitutable) products or services more cheaply than a rival worker, firm, or country. Although these three realms of competition are similar in the abstract, they compete to provide that value in different ways. This post looks at how workers compete.


Workers compete against other workers but they also compete against technologies, such as machines or more efficient processes. In other words, workers are competing against substitutes for the labor they provide, whether that means their neighbors, robots, or their department being made redundant after a merger. As part of the overall production process, however, they further compete with firm owners and managers for the value that the firm receives in exachange for its production.

Firms want workers that will provide surplus value. Firms in turn sell to consumers or other firms. Economic models assume that workers are paid the wage that is equivalent to the value workers provide for the firm, but of course firms would not hire workers if that were exactly true: workers have to provide more value than they are paid or else there is no point in hiring them. The amount of the "surplus" value that workers are able to capture depends on their bargaining power, and worker bargaining power depends on the availability of substitutes for labor and the demand for pay. A worker may accept less pay if it knows a robot could do its job more cheaply. A business will have to pay more if its potential workforce can easily find better pay elsewhere.

Do we overlook the same points about worker competition that we do about other types of competition? Economics does pay attention to changes in demand for workers, and to the pay that workers receive in exchange for filling that demand. Changes in labor demand can be seen through surveys of firm hiring, changes in wages, or even instruments as broad as GDP growth. All these measurements have been linked to the labor markets through extensive study. For example, it is something of a stylized fact in economics that periods of significant overall growth in the economy are the only times when wages increase for the bottom half of the income distribution--this is one common rationale for the gospel of growth. Within-industry trends are also well-studied. It is common to hear where new jobs are expected to be created and which old jobs will be destroyed.

Despite the attention that economists and the general public pay to changing worker incomes, competition is can be hard to understand in aggregate terms. A single unemployed worker might know that competition for work in their occupation is harsh, but it can be hard to see the structural and macroeconomic causes. Economists might see the effects of a bad economy, but the complex array of causes means that they may have little idea why things are bad. For example, experts are still arguing over whether the current high level of unemployment is structural or cyclical.

The idea that competitors shape the size of the pie they are competing for is perhaps most obvious when looking at individual workers, because workers are the ones actually engaging in production. If workers work hard or if they are well trained or just good at their job, they will produce more. If they are unemployed, they will produce nothing. But what determines how productive those workers are?

Competition between workers can in theory help increase productivity: if I will lose my job to someone else for not picking enough tomatoes, then I will pick more tomatoes. I might also go to school and learn things that make me productivity in order to get a better job, and if lots of people do this they may increase overall economic output significantly. Or I might work hard so I can get promoted so that my friends will respect me.

But there is also good evidence that many of the ways we compete --in particular, for money-- are demotivating and could therefore be hurting productivity. Inequality is another possible demotivator, if we don't think we have a fair shot at "making it". My earlier posts on opportunity address this issue from a slightly different angle.

Where competition in labor markets really gets interesting is when we think about whether or not the "size of the pie" is artificially constrained. We see this in the phenomenon of unemployment. Workers compete for jobs but if there are no jobs, on some level it is absurd that there is no work to be done. Of course there is work to be done--but for some reason nobody wants to pay these unemployed people to do it.

But is the opportunity for employment constrained artificially? What would that mean, exactly? Presumably it would mean restrictions on the ability of workers to produce value and receive value in return. The problem is how broadly the the word "restrictions" should be defined. We could imagine restrictions being anything from overtime laws to antitrust laws to public education: overtime laws because some workers may only be able to work a certain job if they are able to gain extra pay working overtime; antitrust laws because smaller producers may be less efficient and that inefficiency may be due to employing more workers; education because workers are only able to provide certain types of value if they have certain training, knowledge or skills.

These are just examples, but it should be apparent that the size of the pie--that is, the total output of all workers--can be constrained both intentionally and unintentionally. This flies in the face of the "lump of labor fallacy", which (controversially!) asserts that demand for labor increases in lock step as the availability of labor increases. For the lump of labor fallacy to truly be a fallacy, you have to assume that firms employ the available workforce fairly and effectively, that supply and demand consistently and quickly match each other. But in reality often the labor supply is extremely "lumpy", as businesses do not hire enough workers to employ everyone who wants a job for many different reasons.

Moreover, we have to remember that the labor market is shaped by laws, relationships, customs, technologies, and a million other things besides an ideally-responsive supply and demand equilibrium. We can do things to help that equilibrium come into existence, that harness competitive forces between workers, between workers and machines, and between workers and firm owners. But the how of it is never as easy as "unleashing competitive market forces" because those forces are only a byproduct of political, technological, and social forces. If we are not careful about where our competition is leading us there is no guarantee it will take us to a world we want to live in. Competition is not that simple.

May 3, 2013

Competition in Context: Firms

In last week's post I outlined five points about competition that tend to go overlooked in economic discourse:
  • The amount of things we are competing for changes over time.
  • The future amounts of things we are competing for are uncertain.
  • The things we are competing for may be created by the competitors themselves.
  • The amount of things we are competing for may be artificially constrained (this is basically a roundabout way of describing the way we often create markets using public or non-rivalrous goods).
  • The current allocation of things we are competing for may have important effects on the ability of competitors to compete for more of them (this is basically a roundabout way of describing things like economies of scale, network effects, path dependency, and first-mover advantage).
I want to take these ideas and look at what they can mean for competition situated in various contexts. What does it mean for a country to be competitive as compared with a competitive business or a competitive worker? I don't want to give a full account of how competition in these areas works, rather to think about why the points I've made above are important.

In all three of these competitive arenas, competitiveness comes from offering more value than competitors are offering. Typically 'offering value' means selling similar (aka substitutable) products or services more cheaply than a rival worker, firm, or country. Although these three realms of competition are similar in the abstract, they compete to provide that value in different ways. Today's post is about competition between businesses.


Firms are more straightforwardly in competition for value than nations are. Unlike nations, there is no exchange rate for firms, so firms don't simply get poorer, they "lose"--broken apart in the messy death of bankruptcy. While nations could be said to compete for value only as a means to an end (such as citizen welfare), firms are competing solely for revenue. And by competing in more or less distinct markets for products and services, the competitive dynamics between firms are easier to see.

Those distinct markets are more volatile than the macro-level "markets" that national economies compete for, so firms face a greater degree of change in size in the markets they are competing in. A population of fish may collapse due to overfishing, or the demand for bottled water may increase exponentially--such changes are rarely so dramatic on a national scale. This uncertainty increases competitiveness between firms because it means that incumbent firms have to adapt to continue to "win".

More specifically, however, the dynamic nature of certain markets can create an interesting variety of competitive dynamics. Competition in growing markets is very different from "mature" or from shrinking markets, because firms can grow even if they might be a worse option than a competitor--witness all of the bad investments that were initially successful in the dotcom bubble. In a shrinking market, on the other hand, even competitive firms can get squeezed down to nothing. (e.g. Kodak in the consumer camera market as it lost ground to camera phones)

While firms compete in ways different from national economies, they are similar in that they are creators of the value (good and services) that they want to exchange for other value (revenue they then reinvest or pay to shareholders). As individual firms attempt to "build up" value by attracting it away from their competitors, this can increase the size of the overall pie that they are competing for. But it does not have to, as we will see below.

Firms build up value by organizing production and utilizing technology in ways that reduce costs; the firms that do this best are rewarded with more money from customers. They are rewarded with less money from each customer--because that is why the customers chose this particular firm's products--but more customers are paying so they have the potential to receive more value overall. By trying to provide the most value for the least in return, we see firms experiencing the same competitive dynamics as when national economies "race to the bottom".

In our previous posts about efficiency and costs, we discussed how some cost reductions are things we would think of as "effiency increases" while others simply shifting costs from one party to another.

This idea helps us see the two possibilities when firms reduce costs: if they truly increase their output per worker they increase the size of the pie that everyone is competing for; but if their cost reductions are based on redirecting value away from the "costs" of workers, purchasers of the product gain exactly the expense of the workers. Alternatively, businesses can inflate the "value" of their products if they are not facing adequate competition. I don't want to get into the issues associated with this pattern of loss/gain here; the point is that because of the fungibility of money, competition can incentivize increasing the size of the pie or it can also incentivize redirecting the value within the pie. And the latter can be problematic.

The final two misconceptions about competition are right on target here. Network effects and artificial scarcity are two important ways that firms can redirect flows of value toward themselves in defiance of what we might think of as an "ideal" market. Any attempt to use market forces for the public good needs to recognize these pitfalls.

April 30, 2013

Competition in Context: National Economies

In last week's post I outlined five points about competition that tend to go overlooked in economic discourse:
  • The amount of things we are competing for changes over time.
  • The future amounts of things we are competing for are uncertain.
  • The things we are competing for may be created by the competitors themselves.
  • The amount of things we are competing for may be artificially constrained (this is basically a roundabout way of describing the way we often create markets using public or non-rivalrous goods).
  • The current allocation of things we are competing for may have important effects on the ability of competitors to compete for more of them (this is basically a roundabout way of describing things like economies of scale, network effects, path dependency, and first-mover advantage).
I want to take these ideas and look at what they can mean for competition situated in various contexts. What does it mean for a country to be competitive as compared with a competitive business or a competitive worker? I don't want to give a full account of how competition in these areas works, rather to think about why the points I've made above are important.

In all three of these competitive arenas, competitiveness comes from offering more value than competitors are offering. Typically 'offering value' means selling similar (aka substitutable) products or services more cheaply than a rival worker, firm, or country. Although these three realms of competition are similar in the abstract, they compete to provide that value in different ways. Today I'll write about competition in national economies.

National Economies

National economies compete with each other in a number of ways, most prominently in trying to attract international investment or run a positive balances of trade (by exporting more final and intermediate goods than they import from other countries). They also compete in other ways, by attracting immigrants or more overtly through war, but I will just consider trade and investment here.

Both investment and trade flows fluctuate significantly over time, and they do so in both net as well as gross terms. This fluctuation--the resizing of the pie the economies are competing for--can have important macroeconomic consequences for countries. These kind of changes in trade and investment are well-enough understood by economists, but they can be missing from popular rhetoric in important ways. For example, protectionist policies during the great depression are thought to have exacerbated the economic downturn, but were still pursued because lawmakers (evidently) failed to see how decreasing US imports could also eventually decrease US exports.

It is worth noting that since the 1970s competition between nations has had a sort of "equalizing" mechanism that should serve to dampen competition: free-floating exchange rates. In theory this means that nations are unable to trade their way into a larger stock of value--they are only able to trade for the value of what they are producing (or can convince other countries they will produce in the future). In basic neoclassical economic theory, therefore, countries are not "in competition" in the way that firms or workers are. In practice, however, trade flows can be heavily dependent on skewed terms of value exchange if exchange rates do not adjust flawlessly (or exist, in the case of Europe).

National economies also reflect the uncertainty of competition (e.g. nations may embark on ambitious industrial planning that fails to come to fruition in the long term) and the "self-determining pie" nature of the value being competed for (e.g. nations may fail to create enough domestic demand to grow internally without relying on trade export surpluses, like China). These misconceptions about competition can have important consequences, like wasted public investment or beggar-thy-neighbor pursuit of competitiveness that ignores the importance of domestic demand. Attempts to promote competitiveness need to reflect the complexities of the real world.

Finally, competition between national economies is also powerfully affected by network effects, path dependency, institutional strength, and other forces acting on markets outside of normal supply/demand issues. This has been recognized by economics but it has been a struggle to put it into practice for many countries: some countries have been successful, such as the "asian tigers" and more recently the BRICs, but others have been unable to harness these positive externalities for their competitive gain. Economics has incorporated some of these ideas into standard trade models but overall they fail to capture many ideas useful for policy.

April 22, 2013

Production as Privilege: Thoughts on Competition

I started writing a post about mercantilism, but I realized I needed to do a bit more thinking about competition first. The last post on this blog was talking specifically about the effect of competition on how people consume, but this is looking more generally at the nature of competition within market. This post is part of a series called Production as Privilege, looking at the way that production relates to the distribution of wealth.

Conceptual Tool #11: Understanding Competition

Competition is a fundamental concept of economics but it is usually restricted to a simplistic market framework of buyers and sellers, or game theoretical choices and payoffs. Let's think a bit harder about what competition is: what we are competing for, who is doing the competing, what are we competing with, etc...

To start with, here are two basics we should be able to agree on:

Competition requires at least two entities. Usually these are firms or individuals, but we can also think about competition between other aggregations: classes, industries, cities or countries.

Competition must be for something that is limited in some way. In economics these are somewhat unhelpfully known as "rivalrous" goods. The point is that they can run out, and each competitor can't have as much as it wants. We don't have to compete for air unless we are underwater.

There are several ways in which we think about competition that are misleading.

First of all, we tend to think about it in static terms, with competitors gunning for a set quantity of stuff. This is rarely true: competitors are often chasing after expanding or contracting amounts of stuff--where "stuff" can be anything from wind-generating capacity to customers who want to purchase typewriters. People have recognized this dynamic element of competition and it gets reflected in business strategies like "blue ocean" strategy.

Once we recognize the dynamic nature of competition, it becomes clear that competition has a probabilistic component as well. We often compete to optimize our returns in the future, and such returns necessarily involve uncertainty. We compete for uncertain outcomes when we buy raffle tickets that might win us a new TV, or when we create export-processing zones that we hope will attract foreign investment.

The third thing we often fail to recognize, and this is perhaps the most interesting point I'm making, is the way that competitors' own actions can shape the availability of what they are competing for. That is, competitors may be competing for pieces of the pie, but the pie may expand or contract based on the actions of the competitors. This is because the fungibility of money converts many different types of competition into competition for value, and we produce value in order to exchange it for other value. This is not a trivial point: it is the crux of Adam Smith's (and all modern free trade supporters') argument against mercantilism (see here) and for the productive power of self-interest.

Fourth, we also forget that much of competition is for things that are artificially scarce. Certainly most newer digital products are kept artificially scarce through copyrights and other legal protections. What's less commonly understood is that compound interest on loans creates scarcity as well, by asking people to pay back more than their original amount. Some bankruptcies are thus actually required because not everyone can win. (Although it's also true that new loans can be created; so I'm not sure I fully understand the implications here.)

Finally, while economics is well acquainted with problematic economies of scale and network effects, the extent to which there is free entry into markets and truly open competition is often overstated. Mass retail markets are dominated, perhaps by definition, by massive economies of scale. Many primary and non-consumer markets are basically oligopolistic--one rationale for deregulated trade is that national-level oligopolies will be replaced by more efficient international competition. And digital markets have entirely different economies of scale than the world has ever seen, where the tools of production and maintaining scarcity take almost any form we can imagine. Here is a fun TEDx talk highlighting network effects and describing the way we misunderstand them when thinking about economics.

When we put all these ideas together we get a picture of competition that doesn't look much like the basic perfect competition/monopolistic competition/oligopoly/monopoly models. Instead there is a dynamic ecosystem of not only firms and workers, but also financers and consumer preferences, and an infinite variety of markets that defy easy classification. Economists know this, of course--they just don't know a precise, mathematical way to think about it.

April 8, 2013

Competitive Consumption

Economics loves competition. It's what makes the economy go 'round. Without it, markets would fall flat on their faces and buyers and sellers would have their arms twisted into horrendous deals. We'd have to pay even more for our mobile phone service than Canadians do.

Most of the attention paid to competition, though, is about competition among firms. The basic types of market competition in econ101 are types of supplier competition. We worry about monopoly often enough, but who ever lost sleep about monopsony in a consumer market? (the standard monopsony example is a single employer in a labor market)

On the consumer side, it's just assumed that there will be many consumers and they will all want to pay as little as possible. These assumptions are generally true, of course, but they oversimplify things because they gloss over why we consume.

Economics 101 starts with the idea that we have unlimited wants, and then adds on some qualifying assumptions about how we prefer variety and get diminishing marginal utility from any single type of good. The idea of unlimited wants justifies not only what happens in the economy, but how we understand the whole enterprise of economic theorizing. They rationalize endless growth and the focus on consumption and tell us what economic policy is supposed to do: meet our unlimited desires with scarce resources.

This is handy logic, but it only takes you so far. You end up with a lot of consumers who want more of certain things because... well, because they want more of them. Metaphorically, we can think of this model as the "sky's the limit" model of consumption: more is always better. (And if something does start giving you decreasing marginal utility, it just means you want more of something else.) This is the common logic used by economists, but there is less explanatory power here than there might first appear. It does not explain, for example, why happiness does not increase beyond a certain income threshold.

How else are we to understand the motive behind our desires, then? The idea I'll call competitive consumption has some interesting explanatory power, and it also has some far-reaching implications for economic theory.

Competitive consumption is socially-motivated consumption, in a somewhat simplistic want-to-have-more-than-the-other-guy manner. We can also think of it as relative or comparative consumption, or consumption motivated by inequality. Inequality is a powerful motivator, as the right likes to argue. It doesn't have to be somebody getting paid 483 times more than I do, though--just a dollar an hour is enough to drive me into paroxysms of jealousy.

Standard economic theory has developed tools to represent competitive consumption, through ideas like positional goods and prestige markets. But these ideas are seen as exceptions to the norm, when in fact they can be fundamental forces in wide range of markets.

Positional goods are things that we buy because they are scarce--often artificially scarce--and their being scarce confers social status on us. Some goods, like loaves of bread perhaps, are not really positional goods, but most things have at least a positional component. Some are literally positional: a penthouse apartment or a front-row seat. Others are positional because the required expenditure of real resources keeps them scarce, such as weekend ski-trips to Vail or sailing around in giant yachts. Still others are positional because they are kept scarce artificially, like diamonds. Or, like spots in an elite school, by means of reputation and concentration of resources and attention. ?????

Most goods have a positional aspect because that aspect is what motivates, to a large degree, our consumption. Better position is often tied to a real increase in "use value", although economics has as good as given up trying to distinguish what that means. More expensive cars might drive better or more expensive fashion might look better, but these goods are also imbued with social value and that is a powerful reason they are consumed.

Mere quantity can have a powerful impact on positionality as well. Having two or three or four cars can be handy, but it can also show status. Of course, people don't go around saying "I can't believe Rupert only has one car! What a low-class person!" But your kid might hate you if you can't take him to all his after-school activities because your spouse drove the car to work, and the other parents might think you are not a good parent. Use value has its own social implications.

Producers are quite good at exploiting consumers' competitive urges and are able to extract consumer value by providing various "qualities" of a different product that are available at different prices. Alcohol is a perfect example, with quality increasing almost imperceptibly as prices skyrocket. Countless brands also have clearly differentiated models (of cards, computers, headphones, suits, etc...) that are less a reflection of higher costs of production than of the nice infinite ladder of consumer prestige. No matter what your budget is, you can spend it all--and it will never be enough. Part of the reason is simply because we have an extremely wacky (and malleable) idea of what things are worth, but many expensive things have inherent value purely through their expensiveness.

What does it mean to take positional goods seriously and really look at the implications of competitive consumption?

First of all, competitive consumption doesn't mean that consumers are competing to buy a given good--it means that consumers consume for competitive reasons. An extreme conclusion of this idea is that demand can be infinite without increasing welfare. We can consume and consume but our utility, on the aggregate, cannot improve if utility is based on our relative position.

Second, static equilibrium analysis seems woefully inadequate. If goods are positional, then the demand for a good can (and probably does) change every time it is purchased. That is, my purchasing a new BMW might raise the overall demand for BMWs (or lower it, more likely). This is Keynes's "animal spirits" writ large across the entire consumer economy. For marketers this idea may be common sense, but economists rarely take it seriously.

But if we do try to do static analysis, what might our market even look like? A demand curve for a given good is supposed to represent the aggregation of different individuals' "willingnesses to pay", with a few people willing to pay a lot for a good and a lot willing to pay a little for it. So it slopes downward:

Competitive consumption, however, could drive consumers to create that downward slope simply for the sake of having a downward slope. That is, with no variation in preference or anything to do with willingness, demand would still sort itself along a curve reflecting little more than the ability to pay. Everybody wants to be at the top, but they can only get as far as their incomes will take them. (Note that this is looking at demand in the aggregate; assuming consumers have varied preferences, then demand with individual product markets may still be differentiated based on how much consumers want a product. But if we look at all products together we can assume people simply want "more".)

This has important implications for social welfare analysis. Economists generally just look at the area under the curve and call that the social welfare effect. (See the wikipedia article on deadweight loss for some nice graphs.) There are already powerful critiques against this type of analysis (Brad DeLong's "non socratic dialog" is one of my favorites), but if the downward slope of the demand curve is simply an artifact of consumer competitiveness, it leaves welfare analysis on even flimsier footing. If one person's utility is another person's disutility, consumption adds no value.

As mentioned above, the constructed nature of the demand curve is not a passive phenomenon. Marketers understand the competitive nature of consumption and are happy to oblige and encourage consumers' pursuit of utility in a zero-sum game.

Competitive consumption is not the only reason for consumption, but to the extent that it does reflect real motives for consumption it can have real implications for economic theory. And thus also for economic policy. Taking competitive consumption seriously requires us to rethink our ideas about equality and welfare, and the effect of consumption on happiness and aggregate utility. We have built our thinking and our policies on assumptions about human psychology and welfare that are suspect, and we will have a difficult time escaping from our current growth patterns without a better understanding of who we are and what we need. One way to do that is to take competitive consumption seriously.

April 2, 2013

The Good, The Bad, the Optimally Efficient in a Given Environment Subject to Change

This is a guest post today from our friend "Blue". Enjoy!

Economists are often accused of physics envy. This is likely valid. Often the alternative suggested is meteorology, implying a limited capacity to forecast the future. That may be a useful perspective as well, but I’d like to suggest a better, and what I consider more realistic, academic brother: genetics.

Like the economy, genetics is concerned with the survival of entities in a competitive environment. They both attempt to explain the behavior of those entities in terms of what is efficient or useful for their success. They both use lots of statistics and game theory, so at minimum you know they are fun.

Let’s back up, dear reader, and discuss the motivation for this discourse. Many economists believe that what economists are doing is describing inherent laws about the way exchanges work. The Austrians often call this “praxeology” which they roughly define as the deductive human behavior in exchanges (for “exchanges” think big markets where people are buying and selling stuff). Though they will say otherwise, the “human” part is not that all important. Rather the exchange itself gives rise to necessary actions a human must take. Thus the study of economics is, in their view, a study about the immutable characteristics of these markets, the laws that govern them and so on. We see when these exchanges/markets are functioning efficiently and say ‘this is good.’ Functioning well is good. The naturalness of it is right, it to behave as we are naturally intended to. While the Austrians are the easiest to pin down here, many, and in a way most, economists are sympathetic to this view, subconsciously if not explicitly. This is the physics envy, the aspiration of mathematical purity, the dogma.

Then there are the economists who say “whoa there buddy, ol’ pal, maybe it is not so immutable after all.” They suggest that perhaps the mathematical elegance of these endeavors should be put to the test. Let’s see how all that deduction matches up with the data, which is ultimately how physicists test their theories anyways. Of course it turns out most if not all of the theory is just awful at describing the real world. Capital flowing from poor countries to rich? Deregulation leading to more stability?

And these economists say, “well, maybe the immutable aspirations are a little too high, let’s just talk about what’s going on in the short run.” Here the models and theory are a little more useful. They seem to tell us how things work when everything is going fine, but of course they never see the crisis until it is upon us. Like meteorologists we know some immutable characteristics of what we are talking about, much as they know how high and low pressure systems will affect weather patterns, but when any specific storm will come is harder to say. This, to some extent is a legitimate aspiration for economists. Like detecting low pressure air, precipitation, temperatures, macroeconomists can say “ah, yes, look at all that accumulating debt, that hyper-inflated stock market and the exchange rate risk, a crisis is likely.” This seems to me useful. Though ultimately more art than science, it may at least have some claim to legitimacy. What is good then is nothing more than nice weather. We don’t want storms, we want pleasant skies and maybe the occasional breeze for kite flying.

But now, I’d like to turn to my proposal. Genetics. Why is this a better or more appropriate aspiration for economists? Let me first persuade you of its relevance. Both studies, economist and genetics, concern themselves with the dynamics involved in the competition of many heterogeneous entities in a given environment. Yes, genes look to replicate and businesses look to accrue profits, but exchange babies for wealth and we are starting to get there. Funny that both genes and socially constructed business strive for survival, immortality really, perhaps there is something philosophical to be said there, but I digress.

And what, then, if anything, can genetics teach economics? In genetics what is good in one environment may spell disaster in another. The march of evolution is not necessarily progress as much as it is adaptation. It’s contingent, conditional, contextual, the opposite of immutable. Indeed what evolution fosters is not necessarily good or bad; it simply creates utility at a given time and in a given environment. What is natural has no sanctity, it will save your life in the forest only to crush you on the grass lands. What is natural is neither good nor bad, it is simply happenstance, the accrual of millions of interactions between entities, their environment and randomness.

And so then the deep questions, the first principles of economics, are open once again for debate. What does a good economy look like? Just because something is good for an individual, does that necessarily mean it is good for society? What are meaningful metrics for these things (the average lifespan of a small business is 8.5 years)? Can a business live forever? What is similar or different about how humans and businesses interact with the giant, dynamic, evolving, idiosyncratic, instructionally contingent monetary environment we call the economy? Indeed.

March 30, 2013

At the Root of the Matter

The acts of exchange or accumulation are the building blocks from which economics is constructed, but the building blocks themselves contain the relationships of mutuality and domination that lie within, or below, all social life. At the root of the matter lies man, but it is not man the "economic" being, but man the psychological and the social being, which we understand only imperfectly.

Robert L. Heilbroner ~ The Worldly Philosophers (6th Ed.)

March 29, 2013

Price, Pricelessness, and Behavioral Economics

William Poundstone Priceless

William Poundstone ~ Priceless: The Myth of Fair Value (and How to Take Advantage of It)

Priceless is a book about price. It is pop science writing: the author is an author, not a behavioral economist. The pace is quick and most of the chapters, all of which have clever titles, are just a few pages. It eschews most pretense of grand narrative; the bulk of the book is examples of how the "prices that make our world go around are not so solid, immutable, and logically grounded as they appear." But there is enough there to tie these examples, and the stories of the social scientists that studied them, together

Poundstone opens with some eye-catching experiments and real-world examples, and then moves into section on psychophysics. I had never even heard of the discipline, but it turns out to have some revealing insights into the way we think and the way we percieve the world. From Poundstone's explanation, psychophyics is more or less dead today not because it was wrong, but because it basically exhausted its possibility for learning new things about the world.
Psychophysics established that our perceptions are based on a logarythmic (as opposed to linear) scale that, perhaps surprisingly, is the same across different people. If you think something is twice as bright, I think it is twice as bright as well. We may not think it is equally bright to begin with, but the relative change will be roughly the same.

Psychophysics established quite firmly that our perceptions are relative. You have probably seen optical illusions that convinced you something was darker or lighter than it was, based on what it was next to. All of our perceptions are that way: we judge things based on comparisons, not some absolute scale.

Psychophysics also established the phenomenon of "anchoring", which is one of the primary focuses of Priceless. Anchoring is when I change the way you percieve something by showing you something else first: I show you my wallet full of $100 bills and then give you $5, versus I hold up a penny and then give you $5. William Hunt was one of the first to notice the phenomenon and determined it could be influenced by "recency, frequency, intensity, area, duration, and higher-order attributes such as meaningfulness, familiarity, and ego-involvement."

Poundstone links the lessons of psychophysics directly to our perceptions of price. Part Three of the book is essentially a background on behavioral economics, and Poundstone does a nice job of tracing the development of different theories through various psychologists and economists. It's mostly a story of how the idea of an irrational individual came to be accepted by economists, culminating with Prospect Theory and the ultimatum game.

Prospect Theory is an economic theory with a solid technical/quanitative bases, and as such it is described in more mathematical/economics terms over at Wikipedia. Poundstone summarizes it in three points:
  • we use reference points to judge the value of things, so our judgements are fundamentally relativistic;
  • we have significant loss aversion, so we generally would rather have $10 than a 50% chance of having either $0 or $20 (typically we would only take the 50-50 bet if we were offered $30, or where the gain is double the loss);
  • and, there is a "certainty effect", which means that there is a big jump in value from say, 90% to 100% (much greater than, say the difference between receiving $90 or $100).
One implication of prospect theory is that we may not have consistent preferences, an important quality of the rational individual economists prefer to model.

The rest of the book is basically stories of different experiments based on the ultimatum game. The ultimatum game is where one player is told to divide $10 between them and the second player, and the second player can either accept the split or reject it. If the second person accepts the offer, they both get the amount the first person proposed; if the second person rejects the offer, neither player gets anything. There are hundreds of variations of this experiment and many of them are fascinating, but the last half of the book consists of little else besides bite-sized descriptions of these or related experiments. There are also a few chapters that read like an editor was encouraging the author to add some content for people who might want to read Priceless as a manual for pricing.
The book is not, however, anything like a manual. It certainly contains plenty of useful information, but the presentation is more a history of scientists and ideas than a handbook for pricing consultants. In that sense the parenthetical note in the title is a bit misleading, but it wasn't why I was reading the book so I won't complain. Overall Poundstone is an engaging writer and I enjoyed the book quite a bit.


It made me think, though.

There are two ways to understand the behavioral economics that Priceless is talking about. Behavioral economics is sometimes portrayed as proof of a fundamental flaw in more mainstream neoclassical economics, a set of evidence that undermines the most basic premises of rational-individual-based theory. But at other times behavioral economics appears to be just an interesting set of addenda--some lab experiments that help us fine-tune our neoclassical models, and the idea of 'bounded rationality' that advises against assuming too much about how people thing.

Which is it? Fundamental change or just smoothing out some rough patches? And how do we decide? And if it is fundamental change, how do we ensure that the necessary change is understood and acted upon?

Something [prominent behavioral economist] Dan Ariely said in a talk I was listening to the other day may clear things up. Neoclassical economics is fundamentally about examining static situations where many, many things are taken as given. Behavioral economics uses two key insights from other domains of social science to push against these neoclassical assumptions. First, the idea that social forces can have far-reaching impacts on markets, and second the idea that social actors create and alter their environment--including the markets they engage in.

In some situations, ignoring these two ideas may make perfect sense. Consumers do act in rational ways sometimes. But in other situations, social forces and the ability of actors to change their environment can render static equilibrium modeling absurd. Of course, economists make an effort to think about whether their results make sense on the level of individual studies. The problem is that once you have too many assumptions you start to exit the real world, and you may find a significant effect that is in fact insignificant relative to other (usually less measurable) factors.

These studies, realistic or not, tend to accumulate and coagulate into simplified worldviews like "markets are self-correcting". Behavioral economics challenges these views by lending credence, in a gradual way, to complexities that belie common assumptions. They give people the ability to ask, "Why did you assume XYZ when you were trying to explain this when we know things could be X, H, or F?"

Our economy is glued together by prices: they are the information that allocates our resources and drives our labor. But prices can only be as coherent as the people that pay them. We need to pay close attention to both the vagaries of pricing and the limits of what prices can represent. Please forgive the trite ending: there are some things money can't buy.

March 21, 2013

Production as Privilege: Rationalization

This post continues our series examining the connection between production and the distribution of wealth and resources. The series has grown far out of proportion and dragged on far longer than originally intended. But far more exciting than all that is the fact that we have scored an original comic from artist David Yoder. He has a bunch of awesome comics to read or buy on his site, and you should really check them out. With a bit of luck, this will not be the last comic you see here.

Conceptual Tool #10: Rationalization

One of my favorite ways to understand the last few centuries is as a process of rationalization, or as it is sometimes known, "McDonaldization". Rationalization is an important idea both in academia (particularly in sociology) and in actual real-world history. Economics as a discipline, however, has only touched on certain aspects of rationality, and has failed to see it as a connected process.

Rationalization as an academic concept came from Max Weber, who used the Prussian buraucracy as a model. More recently, George Ritzer has repackaged the idea of rationalization in his book The McDonalization of Society, using (you guessed it) McDonald's as the archetype of rationality.

Defining Rationality

So what is rationality? It is a concept ripe for circular definitions--we tend to use it interchangably with "reason", but what does reason mean? What does it mean to be "irrational" or to "rationalize" an idea or a process? Etymologically it is similar to "ratio", which means means to calculate or measure. But when we use the word rationality we are talking calculating or measuring a certain thing: I think the idea hinges on causality.

The best definition I can come up with is: thinking or acting in ways that accurately link cause and effect. By extension this implies that when we act in the world we do so coherently, in ways consistent with our intent and motivation.

In sociology, rationality has a much more specific meaning, which Ritzer describes and I will copy from Wikipedia. There are four components:

1. Efficiency - Choosing the best, quickest, or least difficult means to a given end.
2. Calculability - Emphasis on the quantitative aspects of the product being sold.
3. Predictability - Involves the customer knowing what to expect from a given producer of goods or services.
4. Control - A way to keep a complicated system running smoothly. Rules and regulations that make efficiency, calculability, and predictability possible.
The theme of measurement and calculation is obviously still strong in the sociological definition, and you can understand rationalization as the process of optimizing the function of the organization--that is, making the operation as consistent with intent as possible. Optimization is achieved through the the four components above, all of which are related to measurement in a way, but which together change the organization itself.

In practice, rationalization is a process of breaking down production, and reconstituting it in standard ways (e.g. written down and quantified) that are more successful. Success can be economic or military or anything you can measure or compare. It is not inventing new things; it is inventing new ways to do old things that are more successful beause they are able to achieve the intended ends.

Rationality and Economics

I have been writing about efficiency and its relationship to productivity a good deal lately, so I will not elaborate much here. Suffice to say, efficiency is a well-trodden topic in economics (but it is perhaps more helpful to think in terms of cost reduction). The other three aspects of rationality--calculability, predictability, and control--are concepts economics gives a short shrift.

Calculability is certainly fundamental to economics itself, and something on the mind of every economist, but the idea has a different focus in economics than in sociology. Ritzer is talking about the ability of businesses to quantify aspects of their own business and the world they are interacting with: measuring customers served or the time it takes to fry a burger. Economists think a good deal about what can be quantified and how to do it, but they do not think much about businesses creating quantifiability, which is what rationality is all about.

Similarly, the issue of predictability comes up in economics all the time (particularly in terms of risk and uncertainty for businesses) but this is only tangentially linked to the idea of predictability relevant here. The economic ideas of risk and uncertainty play into business decisions on a micro level, but only so far as the external envinronment influences the business. Ritzer is talking instead about the internal processes of the business and how similar iterations of it can be--iterations over space (McDonald's franchises that serve the same food in the same way) or over time (I want to get the same burger I ordered last time). In economics, these issues typically get lumped into the infuriatingly vague "tastes and preferences" category, or the "value and costs" category which is almost as bad.

Another aspect of predictability, however, is starting to make an impression on economics through the (re-)incorporation of institutional factors into economic models. Predictability is seen as an important feature of market regulation and the provision of public goods. "Personalistic" transactions, based on individual relationships, differ from rational, impersonal transactions based on standardized rules and procedures. Rationalized bureaucracies facilitate well-functioning markets.

Control in the sociological sense crops up in economics primarily through the idea of the "principal-agent" problem, which boils down to the question: "how do principals get agents do do what principals want instead of what agents want?" Principal-agent issues are typically understood as incentive issues, and more control, or control over the right things (pay, bonuses, production targets), is typically seen as the solution. If an economist is looking specifically at intra-firm dynamics, he or she might model interactions using game theory, but in most cases these dynamics are assumed away and firms are assumed to be single units.

In a sense, control can be understood as a component of technology, which economics has famously little understanding of. In years past "technology" was referred to as "technique", which hints at the more human and social aspects: teachers use classroom management "techniques", for example, to control their students.

The difference between control in economics and control in sociology is one of focus. Economics for the most part assumes away control--be it embedded in the firm organization or regulatory structure of a market. Sociology takes a more political view of control and all the ways in which different social forces can influence people's decisions and their views of the world.

How Rationalization Works

Ritzer does not focus on the causal relationships between the different aspects of rationalization (e.g. does predictability cause control or vice versa?). Instead he emphasizes that efficiency, predictability, calculability and control are all necessary, intertwined processes. The processes work together to provide businesses with an assortment of advantages. (Given that businesses are typically controlled by shareholders or owners, however, it may be better to think of these processes as providing shareholders and owners with more advantages.)

To understand the advantages of a rationalized business, simply compare a McDonald's hamburger to a local diner or even a home-cooked meal. Cooking at home is labor intensive--you have to go the store, purchase the meat and buns, start the grill, form the patties, and grill them each yourself. If you are not too good with the grill this can be a stressful process. A local diner is more specialized in that they are making more food at once, and can benefit from the accompanying economies of scale. Where a local diner has more trouble is in ensuring that the customer's expectations are precisely met each time--a different cook on the grill might make burgers differently, or someone just passing through town and eating at the restaurant might not know if the burgers were any good. Nor are local restaurants able to benefit from the economies of scale that franchised restaurants are able to--McDonald's presumably has more clout in the beef market, more money to spend advertising at the Superbowl, and more money to pay taste consultants to formulate the ideal blend of coffee than Bob's Local Diner.

The processes at McDonald's are better established, with established routines for each part of the burger's production. A diner, you could say, does not "produce" a burger. Obviously, not all of these things about the way McDonald's works are always advantages, and a burger on your grill may have certain qualities over one from McDonald's will never match. The point is that there are rational business reasons, given the way our economy is set up and our technology and people's preferences, for a business like McDonald's to exist.

The fast food business is only one example. In government bureaucracies, rationalization also plays an important role. Look at how the government regulates the transportation market. The system is made calculable through hundreds of different measurements: speed limits, point systems for driving infractions, and minimum ages are just a few of the obvious ones. The system is made predictable through clear rules that both drivers and police officers have to follow, and it is made efficient through changes through these laws. Laws also govern control: who can arrest you and who can be arrested; who can control vehicles and must be restrained by a seat belt (everyone).

Market economies based on laws, rights, and regulations must be administered, and the administration works best if it is done by a strong, capable bureaucracy with the correct incentives. While you may scoff at anyone calling the local Department of Motor Vehicles efficient, they do their job: if you fulfill the requirements you will get the license. Success or failure does not depend on whether you know the teller or whether they think you gave them a sufficient bribe. They may piss people off with wait times, but it is a fair system.

Rationalization can help organizations understand and respond to the world they operate in, but the idea is best understood as an internal process that governs actions and relationships within the organization. We can think of entire countries as organizations as well. Constitutional republics like the USA are rational not because they are able to conduct foreign policy that accurately reflects their self interest; they are rational because authority is vested in legal documents (their constitution and laws) and not people. Legal documents are predictable (they say more or less the same thing to anyone who reads them) and create calculable situations by specifying majorities or other criteria for power. The US constitution has many deliberate inefficiencies, but the difficulty in creating or changing laws can reduce other costs--such as unjust imprisonment.

Rationalization, Production and Distribution

Rationalization tends to take the productive process and break it into little pieces, because it is possible to arrange those peices in ways that better achieve the desired goals. The peices themselves might be simplified, but the simplified parts can facilitate a complex whole. (Complexity is not the goal, but a large complex system may outperform small simple ones.) The little pieces are often enhanced by mechanical technology or organizational structures.

The archetypal factory of the industrial revolution, the textile mill, is a perfect example of this. Instead of individual people spinning their own wool from their own sheep and weaving on small looms at home, large factories of workers weave processed cotton ceaselessly on giant steam-powered mechanical looms.

Rationalization is a specific way of understanding the technical improvements and cost reductions I have written about in several previous posts. Once you understand the idea of rationalization you start to see it everywhere, and you start to see how reactions to many other social issues are in a sense reactions to rationalization.

But we are concentrating here on production and distribution. The idea of rationalization is powerful because it starts us thinking about how technological change can remake the productive process--in ways that ultimately shape distribution. What ways am I talking about? Rationalization changes the skills required by workers, often polarizing low and high-skill jobs into lower and higher-skill jobs. It changes, often increasing, the nature of control in workplaces. It increases the scale of businesses. And it lowers costs in a myriad other ways, ways that have a wide range of consequences. These effects of rationalization have been some of the most powerful forces shaping distribution in the past two centuries.

Rationalization can cause skill polarization by reducing the required skill for some jobs while increasing the required skills for others. Think of the skill gap between the McDonald's burger flipper and the McDonald's food scientist with a PhD in Mouth Feel, and how they are the rationalized evolution of your dad standing outside by the grill. Unskilled labor is easily replaceable and cheap, and thus good for the bottom line. This is particularly important because of the scale of operations--the vast majority of employees are near the unskilled end. The skilled worker can command higher prices because they are inherently scarcer (due to things like PhD admissions). The polarization is not symmetrical; rather it ends up being a pyramid without a middle with a large cohort of unskilled work and a smaller tip of skilled labor.

Technology as a general concept may have an ambiguous effect on skills required of workers: it can deskill work or it can eliminate work (depending on the type of technology); it can make workers more productive or less necessary (depending on the demand for the product); and it can increase or decrease the skills required of workers. A cash register might eliminate a worker's need for math skills; but it might also increase the need for a worker to multitask, pouring drinks and assembling orders while answering a customer's question about calorie counts. Similarly, an advanced robot in an automobile factory might simplify employees' jobs to the push of a button, or it might require them, or a quarter of them, to know advanced robotic programming.

But technology in the context of rationalization is less ambiguous. As noted above, control itself can be understood as a technology: the processes rationalization touches are deliberately simplified, dumbed down and subjected to more control--partly because control enhances predictability, partly for its own sake, and partly because quantified, simplified processes make control easier. Although I do not have data on this, I would expect the overall trend within industries to be that of skill polarization.

Control can impact distribution of wealth and resources in ways that are not well understood by economics. Economics focuses on markets, but markets can only provide limited accountability in terms of price competition and demand. Control requires either hierarchy or some other accountability structure, which is why firms essentially shelter their internal operations from the market. As an example, pay is an important part of establishing hierarchy, and pay may be linked more closely to status in the hierarchy than with required effort, skill, or even supply and demand.

Rationalization also tends to shift control not to people but to written rules and procedures, shrinking the amount of people who actually shape the institution. Managers may have little to do with the actual design and processes of the business, instead simply enforcing what has been written. This standardization can have enormous benefits in terms of productivity and the ability of the organization to adapt, grow, and succeed. But the powerlessness of people within an organization can also be problematic not only in term of wages but in terms of rationality itself: anyone who has ever worked for a large organization quickly realizes there are some things that do not make sense but are nevertheless done "because that's just how we do things around here."

A third fundamental way in which rationalization impacts the production and distribution is through increasing scale. As somebody once said (apparently this is attributed to Stalin but no one is sure who said it first), "Quantity has a quality all its own." Scale changes many things, but it is not immediately clear what.

Scale in business has the tendency to concentrate profits near the top because that is basically the point of the corporate model of control. If you have 100 different independently owned burger restaurants, those profits stay in each restaurant. But with McDonald's, a portion of those profits go back to corporate headquarters and then on to investors. McDonald's is a franchise, so in some ways the scale is actually not as large as other corporate entities. But the point is that at McDonald's, operations are standardized in cost-minimizing ways, ways that facilitate the minimization of employee wages across the board.

Assuming that there are certain market sectors that become crowded with competitive entrants, we can also think of scale as prohibiting new experimentation and diversity. A national market for fast food or groceries or auto parts or anything controlled by two or three huge players is a market that is difficult for small players to enter and compete within. Larger businesses have more leverage in the labor marketplace (as well as many other markets) and are less likely to be susceptible to institutions that increase worker's ability to gain a part of the profits. For example, if the person who controls a basic wage rate for someone flipping burgers is far away at corporate headquarters, there may be less pressure to pay a living wage than if the business owner lives down the street.

In the end rationalization is primarily about reducing costs in the above-mentioned ways. Scaling up, increasing control, and decreasing the skill required of workers all help lower costs to organizations. This can be a very good thing: if cars were built by hand by a single craftsman nobody would have a car. But costs can also be a zero-sum game. Institutions that have become too rational can end up being irrational for a large portion of the people they affect, because their rationality is too narrowly considered. This is a major point of Ritzer's. Rationalization may come with costly externalities, like pollution or repetitive stress injuries, or it may simply shift costs in ways that benefit some at the expense of others.

Rationalization is not going away; if anything it is only in its infancy. Organizations like Samasource specialize in breaking down work even further, rationalizing processes that were not previously economically feasible. Control over employees is also evolving, as new technologies help quantify the previously unquantifiable. And technology continues to encroach on scarce abilities.

As our technologies, our workplaces and our work change, we need to use all the tools at our disposal to understand the economic as well as the social impact of those changes. The concept of rationalization helps reveal the how the forces shaping businesses can have social impacts as well as distributional consequences.