In this article I'm going to try to explain, using very stylised facts, why America grew so delightfully through the late 19th to late 20th centuries, and why it has seemingly hit a roadblock recently, with implications for OECD development policy going forward. I think America's story nicely illustrates some changes in the basic structure of advanced economies that have taken place in the last 20 years that have substantial policy implications, particular where the role of governments and markets is concerned. What I'm writing here is embryonic in my own thinking, but I think it's articulate enough to post.
From basic microeconomic and growth theory, we would expect America to grow very well in the 20th century because of its liberal institutions and policy settings - what Acemolgu and Robinson call, perhaps inaccurately, 'inclusive' institutions. These are institutions that allow everyone relatively equal access to the economy. Everyone is 'included' in the economy. Such institutions are juxtaposed with extractive institutions (like those of colonialism and feudalism), where some people's economic opportunities are deliberately limited to ensure the power and wealth of a minority elite. Inclusive institutions provide, in a sense, a fairly high degree of equality of opportunity, at least by the standards of a developing country, which is what America was at the time in terms of GDP/capita (by this I mean, for example, that almost nobody had access to high levels of education or health in America in 1900). Property rights and contract law were well enforced, the government imposed very few distortions on markets in the form of price floors, price ceilings, subsidies and taxes, and what regulation did exist was largely to prevent anticompetitive behaviour rather than to promote equity by way of market interventions. As such, it was relatively easy to start a business and grow it. Moreover, America had an enormous geographic area to develop. This was a powerful combination. Factor inputs (labour and capital) were very mobile and could go where productivity (and thus rents and wages) were highest. Moreover, these inputs were attracted to America from abroad (migrants especially) by the promise of land and economic freedom.
What you had in America over the course of the 20th century was basically institutional and policy settings that lead to classic market efficiency as explained in microeconomics 2. Economic outcomes are not equitable, but efficiency is very high so the total economic pie gets very large, quickly. Job growth is also tremendous, so while some people are worse off than others, everyone feels like their lot is improving. Negative externalities aren't so palpable because the capacity of the environment to absorb them is high owing to its pristine initial state, and the small-scale public goods required at this time are relatively easily provided by local government, which is left to address local conditions by a relatively minimalist state and federal government architecture. Monopoly was checked, at least to an extent, by the antitrust laws passed in 1890 and 1914 to break-up the Robber Barons, and by the huge geographic area of the US, which makes it possible for firms to emerge catering to a market in several different parts of the country before meeting at the borders to compete. These are really the glory years of laissez-faire capitalism, and it is unsurprising that America not only formalised a lot of the principles involved (notably in neoclassical economics, which is largely an American invention) but also became ideologically committed to them.
A critical point to make is that factor-mobility throughout this period is very high in part because no jobs require high-level skills. As a worker, you can get a job at a factory for 20 years (your working life) without much trouble, but you can also move quickly to a new industry if your present employer goes bust because all industries require basically the same skills. Very few industries require particular niche skills and workers can be retrained quickly because everything is basically about your hands and learning processes rather than ideas. For example, you basically have exactly the same skill set for clerks, foremen, agricultural labourers and army officers. Nobody needs more than high school literacy and numeracy. Compare this to today where lawyers, accountants, software engineers, economists etc. all have quite radically different skill sets that take years to learn.
This plugs into work by people like Autor that examines why America's workforce is not bouncing back from the 'China's shock' that saw its manufacturing sector move off-shore (offshoring is a major change to capital mobility that has taken place rapidly over the past 20 years) . Russ Roberts, a libertarian-inclined economist, thinks it must be because rigidities have been imposed on markets in recent years that are preventing the kinds of adjustments that happened smoothly during the 20th century. He is particularly concerned about labour regulations and minimum wages. My own view is that labour mobility in the 21st century is a very different beast to labour mobility in the 20th century and, moreover, that it actually requires a substantial state apparatus to ensure smooth labour mobility nowadays, hence why the US manufacturing workforce seems to have fallen into a hole it can't get out of - there is no state to draw on. I certainly don't think regulations help, but I also think 'smaller government' is absolutely the wrong answer to the current malaise in deindustrialised areas.
Labour mobility seems to me to require 2 things (well 3, but I'll leave the third for the end of this article): the ability to get a job somewhere else, and the ability to move to that job. In the growth years of the 20th century, getting a job in America was pretty straightforward. As a high school drop out you could walk around any major city for a day or two and find work going that suited you. Moving was still expensive, but you could take a loan and be pretty certain you could pay it off shortly after arriving at your new destination.
Not so nowadays. Getting a job often requires substantial skills that aren't taught on the job (because they can't be - you can't, for example, teach someone South-East Asian security analysis on the job). These skills often take a long time to develop, and cost large amounts of money. The skills are also often quite niche, so finding an employer who wants you can be difficult, even with the internet and even if there are jobs going. With the pace of technological change and the geographic fragmentation of industries, it's also hard to know what is a safe choice when retraining.
To me, many of these things seem like either market failures (notably associated with information and transaction costs) that require government correction or examples of areas where the government can play a very helpful (and efficiency-enhancing) role as risk-manager and facilitator. I will address these in turn. The most comprehensive reference I have for this stuff is Slazar-Xirinachs, Nubler and Kozul-Wright (2014), 'Transforming economies: making industrial policy work for growth, jobs and development', ILO and UNCTAD. It's a free ebook that you can find with a google.
First, on market failure, there is a sophisticated literature emerging out of research into industrial policy by Justin Lin and other people in the lefter branch of the World Bank and other apex organisations pertaining to the ways government can effectively and usefully coordinate economic activity. The emphasis in this literature is precisely on reducing information shortfalls and transaction costs. There is a longstanding and very reasonable skepticism of government in this regard that dates back to the original debates around central planning and laissez-faire capitalism from about 1900-1960. Economic coordination in the form of central planning requires enormous amounts of information that a bureaucracy can't possibly compile or analyse. Markets take care of this processing in a diffuse way by tapping local knowledge, notably of perpetually shifting preferences and relative costs. Without this information, government's are not better at picking winners (i.e. recipients of state funding) than markets, and the advent of passive funds has shown that markets as a whole consistently pick winners much better than even some of the most sophisticated individual players in those markets. However, this new literature is not advocating central planning, nor does it require the government to pick winners. Instead, it investigates ways for governments to follow initial market signals, like foreign direct investment into emerging comparative advantages, and provide greater thrust to accelerate transitions. Government's main role is to provide infrastructure and easy business processes (single clearance windows, straightforward regulation etc.) and other public goods that reduce transaction costs and make a nation an attractive business destination, and to act as a channel through which information can pass through. For example, the government can act in a communicating capacity between employees, education and training institutions and small scale investors and large firms looking to come in from abroad. For the large firms to negotiate individually with the other actors would be prohibitively expensive. But the government can represent them all and then activate its bureaucratic networks to spread information across the system. This requires large, professional governments, and it is unsurprising that Denmark is perhaps the best at this among the most developed nations.
Second, on risk management and facilitation, it's abundantly clear that from now on, people in the OECD will need to change jobs frequently throughout their life, often with retraining in between. This kind of lumpiness in behaviour is something that behavioural economics has demonstrated people are very bad at doing in the sense that they do not 'rationally' optimise over long time horizons. They don't save enough, they don't get ahead of the curve, they dislike change etc. So this is one of those areas where I think if you take a neoclassical economic point of view, you're going to end up with terrible social outcomes, which is what I think we're seeing in the US and UK, especially as there the neoclassical point of view as implemented by the Reagan/Thatcher/etc administrations was barely as sophisticated as microeconomics 1. What you need to overcome the cognitive biases in these areas identified by behavioural economics is implement policies and institutions that help people to make immediate decisions at one point in time that do the smoothing-over-a-long-time-horizon for them. There are many ways of doing this. The original 'learn-fare' dimensions of Denmark's flexicurity system is a great example, as is income contingent loans for retraining and migration. Something like levy-financed free higher education would work as well. The US has none of these things, and the UK isn't much better, but at least they have the NHS. The government here, which is eternal, uses the magic of inflation-indexed interest rates to smooth risk, consumption and investment for citizens over the life course at next to no explicit cost and with implicit benefits to equity and economic efficiency. The government has an enormous comparative advantage in this sort of thing, in part because of what Stiglitz calls transactional efficiency, much of which comes out of the capacity of the tax system to manage repayment in an automated way.
I think this is the direction that the mythical 'alternative to capitalism' is going to come from (with a lot of support from the green movement and new technologies like the internet that boost productivity in ways that are largely unaccounted for in GDP), but really it's just an update to the welfare state. It's a third way redux.
One angle on all this that I don't think has been considered, especially by global liberalists (like myself) even though all their models depend on it, is that the third thing labour mobility requires is a psychic willingness to move. This is a big part of why some right-wing types resent welfare payments - they think they discourage people from moving to where there is work, and there is a lot of truth to that. One angle that I want to bring up is that there is still heaps of well-paid work available internationally for people with skills pertaining to manufacturing that were acquired in 1st world production lines. However, these jobs are no longer geographically proximate - they're in ASEAN, central America and Eastern Europe. I can't imagine that if you were a manager in the Holden factories of Australia that you couldn't get a similarly paid (because of cost of living reductions) job in the new Malaysian operation, you just need to move to Malaysia, and that's not something people want to do. I don't think the international economic policy technocracy takes this dimension of welfare very seriously (in part because they are very happy to move for their own jobs), but I'll leave a discussion of this to another day.
From basic microeconomic and growth theory, we would expect America to grow very well in the 20th century because of its liberal institutions and policy settings - what Acemolgu and Robinson call, perhaps inaccurately, 'inclusive' institutions. These are institutions that allow everyone relatively equal access to the economy. Everyone is 'included' in the economy. Such institutions are juxtaposed with extractive institutions (like those of colonialism and feudalism), where some people's economic opportunities are deliberately limited to ensure the power and wealth of a minority elite. Inclusive institutions provide, in a sense, a fairly high degree of equality of opportunity, at least by the standards of a developing country, which is what America was at the time in terms of GDP/capita (by this I mean, for example, that almost nobody had access to high levels of education or health in America in 1900). Property rights and contract law were well enforced, the government imposed very few distortions on markets in the form of price floors, price ceilings, subsidies and taxes, and what regulation did exist was largely to prevent anticompetitive behaviour rather than to promote equity by way of market interventions. As such, it was relatively easy to start a business and grow it. Moreover, America had an enormous geographic area to develop. This was a powerful combination. Factor inputs (labour and capital) were very mobile and could go where productivity (and thus rents and wages) were highest. Moreover, these inputs were attracted to America from abroad (migrants especially) by the promise of land and economic freedom.
What you had in America over the course of the 20th century was basically institutional and policy settings that lead to classic market efficiency as explained in microeconomics 2. Economic outcomes are not equitable, but efficiency is very high so the total economic pie gets very large, quickly. Job growth is also tremendous, so while some people are worse off than others, everyone feels like their lot is improving. Negative externalities aren't so palpable because the capacity of the environment to absorb them is high owing to its pristine initial state, and the small-scale public goods required at this time are relatively easily provided by local government, which is left to address local conditions by a relatively minimalist state and federal government architecture. Monopoly was checked, at least to an extent, by the antitrust laws passed in 1890 and 1914 to break-up the Robber Barons, and by the huge geographic area of the US, which makes it possible for firms to emerge catering to a market in several different parts of the country before meeting at the borders to compete. These are really the glory years of laissez-faire capitalism, and it is unsurprising that America not only formalised a lot of the principles involved (notably in neoclassical economics, which is largely an American invention) but also became ideologically committed to them.
A critical point to make is that factor-mobility throughout this period is very high in part because no jobs require high-level skills. As a worker, you can get a job at a factory for 20 years (your working life) without much trouble, but you can also move quickly to a new industry if your present employer goes bust because all industries require basically the same skills. Very few industries require particular niche skills and workers can be retrained quickly because everything is basically about your hands and learning processes rather than ideas. For example, you basically have exactly the same skill set for clerks, foremen, agricultural labourers and army officers. Nobody needs more than high school literacy and numeracy. Compare this to today where lawyers, accountants, software engineers, economists etc. all have quite radically different skill sets that take years to learn.
This plugs into work by people like Autor that examines why America's workforce is not bouncing back from the 'China's shock' that saw its manufacturing sector move off-shore (offshoring is a major change to capital mobility that has taken place rapidly over the past 20 years) . Russ Roberts, a libertarian-inclined economist, thinks it must be because rigidities have been imposed on markets in recent years that are preventing the kinds of adjustments that happened smoothly during the 20th century. He is particularly concerned about labour regulations and minimum wages. My own view is that labour mobility in the 21st century is a very different beast to labour mobility in the 20th century and, moreover, that it actually requires a substantial state apparatus to ensure smooth labour mobility nowadays, hence why the US manufacturing workforce seems to have fallen into a hole it can't get out of - there is no state to draw on. I certainly don't think regulations help, but I also think 'smaller government' is absolutely the wrong answer to the current malaise in deindustrialised areas.
Labour mobility seems to me to require 2 things (well 3, but I'll leave the third for the end of this article): the ability to get a job somewhere else, and the ability to move to that job. In the growth years of the 20th century, getting a job in America was pretty straightforward. As a high school drop out you could walk around any major city for a day or two and find work going that suited you. Moving was still expensive, but you could take a loan and be pretty certain you could pay it off shortly after arriving at your new destination.
Not so nowadays. Getting a job often requires substantial skills that aren't taught on the job (because they can't be - you can't, for example, teach someone South-East Asian security analysis on the job). These skills often take a long time to develop, and cost large amounts of money. The skills are also often quite niche, so finding an employer who wants you can be difficult, even with the internet and even if there are jobs going. With the pace of technological change and the geographic fragmentation of industries, it's also hard to know what is a safe choice when retraining.
To me, many of these things seem like either market failures (notably associated with information and transaction costs) that require government correction or examples of areas where the government can play a very helpful (and efficiency-enhancing) role as risk-manager and facilitator. I will address these in turn. The most comprehensive reference I have for this stuff is Slazar-Xirinachs, Nubler and Kozul-Wright (2014), 'Transforming economies: making industrial policy work for growth, jobs and development', ILO and UNCTAD. It's a free ebook that you can find with a google.
First, on market failure, there is a sophisticated literature emerging out of research into industrial policy by Justin Lin and other people in the lefter branch of the World Bank and other apex organisations pertaining to the ways government can effectively and usefully coordinate economic activity. The emphasis in this literature is precisely on reducing information shortfalls and transaction costs. There is a longstanding and very reasonable skepticism of government in this regard that dates back to the original debates around central planning and laissez-faire capitalism from about 1900-1960. Economic coordination in the form of central planning requires enormous amounts of information that a bureaucracy can't possibly compile or analyse. Markets take care of this processing in a diffuse way by tapping local knowledge, notably of perpetually shifting preferences and relative costs. Without this information, government's are not better at picking winners (i.e. recipients of state funding) than markets, and the advent of passive funds has shown that markets as a whole consistently pick winners much better than even some of the most sophisticated individual players in those markets. However, this new literature is not advocating central planning, nor does it require the government to pick winners. Instead, it investigates ways for governments to follow initial market signals, like foreign direct investment into emerging comparative advantages, and provide greater thrust to accelerate transitions. Government's main role is to provide infrastructure and easy business processes (single clearance windows, straightforward regulation etc.) and other public goods that reduce transaction costs and make a nation an attractive business destination, and to act as a channel through which information can pass through. For example, the government can act in a communicating capacity between employees, education and training institutions and small scale investors and large firms looking to come in from abroad. For the large firms to negotiate individually with the other actors would be prohibitively expensive. But the government can represent them all and then activate its bureaucratic networks to spread information across the system. This requires large, professional governments, and it is unsurprising that Denmark is perhaps the best at this among the most developed nations.
Second, on risk management and facilitation, it's abundantly clear that from now on, people in the OECD will need to change jobs frequently throughout their life, often with retraining in between. This kind of lumpiness in behaviour is something that behavioural economics has demonstrated people are very bad at doing in the sense that they do not 'rationally' optimise over long time horizons. They don't save enough, they don't get ahead of the curve, they dislike change etc. So this is one of those areas where I think if you take a neoclassical economic point of view, you're going to end up with terrible social outcomes, which is what I think we're seeing in the US and UK, especially as there the neoclassical point of view as implemented by the Reagan/Thatcher/etc administrations was barely as sophisticated as microeconomics 1. What you need to overcome the cognitive biases in these areas identified by behavioural economics is implement policies and institutions that help people to make immediate decisions at one point in time that do the smoothing-over-a-long-time-horizon for them. There are many ways of doing this. The original 'learn-fare' dimensions of Denmark's flexicurity system is a great example, as is income contingent loans for retraining and migration. Something like levy-financed free higher education would work as well. The US has none of these things, and the UK isn't much better, but at least they have the NHS. The government here, which is eternal, uses the magic of inflation-indexed interest rates to smooth risk, consumption and investment for citizens over the life course at next to no explicit cost and with implicit benefits to equity and economic efficiency. The government has an enormous comparative advantage in this sort of thing, in part because of what Stiglitz calls transactional efficiency, much of which comes out of the capacity of the tax system to manage repayment in an automated way.
I think this is the direction that the mythical 'alternative to capitalism' is going to come from (with a lot of support from the green movement and new technologies like the internet that boost productivity in ways that are largely unaccounted for in GDP), but really it's just an update to the welfare state. It's a third way redux.
One angle on all this that I don't think has been considered, especially by global liberalists (like myself) even though all their models depend on it, is that the third thing labour mobility requires is a psychic willingness to move. This is a big part of why some right-wing types resent welfare payments - they think they discourage people from moving to where there is work, and there is a lot of truth to that. One angle that I want to bring up is that there is still heaps of well-paid work available internationally for people with skills pertaining to manufacturing that were acquired in 1st world production lines. However, these jobs are no longer geographically proximate - they're in ASEAN, central America and Eastern Europe. I can't imagine that if you were a manager in the Holden factories of Australia that you couldn't get a similarly paid (because of cost of living reductions) job in the new Malaysian operation, you just need to move to Malaysia, and that's not something people want to do. I don't think the international economic policy technocracy takes this dimension of welfare very seriously (in part because they are very happy to move for their own jobs), but I'll leave a discussion of this to another day.
Comments
Post a Comment