Can you manage what you can’t measure?

Disclaimer: I’m not a macroeconomist, but I play one in Principles classes. Feel free to point out my errors in the comments.

Economics is filled with imperfect measures of important concepts. GDP, unemployment rates, and price indices all have significant flaws. Measures of entrepreneurship are even worse.

The Chicago Booth IGM Experts’ Panel recently posted an interesting result about the problems with unemployment:

IGMChicago_EmploymentandtheUSEconomy

The question is: “The concept of “maximum sustainable employment” is well defined enough to be used beneficially in economic policymaking.”

That’s 42 economists asked, 34 with an opinion (including “uncertain”), and about half of the sentiment is positive. And that sounds about right to me. I read this question as “can we engage in macroeconomic policy that doesn’t do long-run harm?” And if I were to describe my opinion probabilistic, it would look like these results. We could probably do “good policy” with a low degree of precision (i.e. useful enough for severe situations) if 1) we magically didn’t have to deal with political squabbling and 2) explicitly limited our goals to the short run.

Respondents agreeing emphasize that it’s “good enough.” But what does that mean? What level of precision and control are we looking at here?

Here’s the basic situation: given the underlying economic reality (everyone’s individual preferences, capabilities, knowledge, etc.) it makes sense to have some limited number of unemployed people at any given time. Unemployment rate is a measure of “how far are we from 0,” not, “how far are we from the ideal level?” That raises the following questions:

  • How accurately can we know what’s going on in the world?
  • How accurately can we know what’s supposed to be going on in the world?
  • How precisely can we affect the world?

The aggregate macroeconomic theory and empirics of the 1960’s were seriously imprecise. Trying to target macroeconomic policy would be like doing eye surgery with a hand grenade. In that sort of world, your best medicine is probably chicken soup and bed rest.

Here in the present we have much better data and computation. And it’s going to get better. Facebook already has data precise enough to track the exact effects of policy down to the individual level.

But think about a precise outcome and imagine what sort of policy would be required to get there. Let’s say our outcome is “get Rick to buy a new car.” Some mix of low interest rates, the right set of subsidies, and ideal circumstances might move things in that direction. But that sort of surgical outcome is just never going to be possible through legitimate macroeconomic policy (and definitely not monetary policy*). The size of the problem just means that Butterfly Effect problems will prevent macroeconomic policy from having household-level precision, even if the data could (in principle) measure the effect.

Less precise outcomes seem plausible (e.g. “get middle class households in the north east to buy 30-60,000 new cars”), but not without making a lot of second order problems. It’s not so much a “middle of the road leads to socialism” situation as a multiplier effect from a convergent series. Push a billion over there, create big distortions, follow those up with medium distortions, and finally let small remaining problems fizzle out on their own. Throw in public choice issues and it seems obvious that the efficiency cost of precise interventions won’t scale up nicely.

We’re now in a world where our theoretical and empirical tools are more precise than our implementation tools. We’re no longer trying to do surgery with a hand grenade. But that doesn’t mean we’ve got a laser scalpel either. It’s more like we’ve got we’ve got laser-like measuring tools but our implementation tools are blunt objects. If we aren’t careful, we’ll spend our time swatting flies with sledge hammers.

All told, it seems possible that we could engage in fairly targeted macroeconomic policy. But as the scale increases, the marginal cost will rise rapidly (maybe less so with the right institutions). Better timing might increase the effectiveness of each dollar spent*, but overall, it seems like things have to be really bad before the government should consider anything too ambitious.

Manage an economy’s economic health using imperfect measures like unemployment rate is about like trying to manage your physical health using heart rate and body mass index figures. Can we do “good enough”? Only to a limited extent. We could stop eating garbage and get some exercise. But there isn’t some magic combination of vitamins and crystals that will stop cancer. Anyone who tells you otherwise is a) trying to sell you something and b) trying to convince themselves they’re in control of the uncontrollable.

In terms of macroeconomic policy, we should start by caring about fundamentals (i.e. long run economic development) and be skeptical of people promising to control the business cycle.


*And let’s face it, we aren’t likely to get decent macroeconomic policy (in terms of efficiency, timing, and appropriateness of policy to circumstances) out of Congress until there’s a serious cultural shift.

Advertisements

Ten best papers/books in economic history of the last decades (part 2)

Yesterday, I published part 1 of what I deemed were the best papers and books in the field of economic history of the last few decades. I posted only the first five and I am now posting the next five.

  • Carlos, Ann M., and Frank D. Lewis. Commerce by a frozen sea: Native Americans and the European fur trade. University of Pennsylvania Press, 2011.

This book is not frequently cited (only 30 cites according to Google Scholar), but it has numerous gems for scholars to include in their future work. The reason for this is that Carlos and Lewis have pushed the frontier of economic history into the history of Natives in the New World. This issue of Natives in North America is one of those topics that irritates me to no end as an economic historian. A large share of the debates on economic growth in the New World have been centered on the idea that there was either some modest growth (less than 0.5% per year in per capita income) or no growth at all (which is still a strong testimonial given that the population exploded). But all that attention centres on comparing “whites” (and slaves) in the New World with everyone in the Old World. In the first decades of the colonies of Canada and the United States, aboriginals clearly outnumbered the new settlers (in Canada, the native population around 1736 was estimated at roughly 20,000 which was slightly less than the population of Quebec – the largest colony). Excluding aboriginals, who comprised such a large share of the population, at the starting point will indubitably affect the path of growth measured thereafter. My “gut feeling” is that anyone who includes natives in GDP accounting will lower the starting point dramatically. That will increase the rate of long-term growth. Additionally, the output that aboriginals provided was non-negligible and probably grew more rapidly than their population (the rising volume of furs exported was much greater than their population growth). This is why Carlos and Lewis’s work is so interesting: because it is essentially the first to assemble economic continuous time series regarding trade between trappers and traders, the beaver population, property rights and living standards of natives. From their work, all that is needed is a few key defensible assumptions in order to include natives inside estimates of living standards. From there, I would not be surprised that most estimates of growth in the North American colonies would be significantly altered and the income levels relative to Europe would also be altered.

  • Floud, Roderick, Robert W. Fogel, Bernard Harris, and Sok Chul Hong. The changing body: Health, nutrition, and human development in the western world since 1700. Cambridge University Press, 2011.

This book is in the list because it is a broad overview of the anthropometric history that has arisen since the 1980s as a result of the work of Robert Fogel. I put this book in the list because the use of anthropometric data allows us to study the multiple facets of living standards. For long, I have been annoyed at the idea of this unidimensional concept of “living standards” often portrayed in the general public (which I am willing to forgive) and the economics profession (which is unforgivable). In life, everything is a trade-off.  A peasant who left the countryside in the 19th century to get higher wages in a city manufacture estimated that the disamenities of the cities were not sufficient to offset wage gains (see notably Jeffrey Williamson’s Coping with City Growth during the British Industrial Revolution on this). For example, cities tended to have higher food prices than rural areas (the advantage of cities was that there were services no one in the countryside could obtain).  Cities were also more prone to epidemics and pollution implied health costs. Taken together, these factors could show up in the biological standard of living, notably on heights. This is known as the “Antebellum puzzle” where the mean heights of individuals in America (and other countries like Canada) fell while there was real income and wage growth. The “Antebellum puzzle” that was unveiled by the work of Fogel and those who followed in his wake represents the image that living standards are not unidimensional. Human development is about more than incomes. Human development is about agency and the ability to choose a path for a better and more satisfying life. However, with agency comes opportunity costs. A choice implies that another path was renounced. In the measurement of living standards, we should never forget the path that was abandoned. Peasants abandoned lower rates of infant mortality, lower overall rates of mortality, the lower levels of crowding and pollution, the lower food prices and the lower crime rates of the countryside in favor of the greater diversity of goods and services, the higher wages, the thicker job market, the less physically demanding jobs and the more secure source of income (although precarious, this was better than the volatile outcomes in farming). This was their trade-off and this is what the anthropometric literature has allowed us to glean. For this alone, this is probably the greatest contribution in the field of economic history of the last decades.

  • De Vries, Jan. The industrious revolution: consumer behavior and the household economy, 1650 to the present. Cambridge University Press, 2008.

Was there an industrious revolution before the industrial revolution? More precisely, did people increase their labour supply during the 17th and 18th centuries which lead to output growth? In proposing this question, de Vries provided a theoretical bridge of major significance between the observations of wage behavior and incomes in Europe during the modern era. For example, while wages seemed to be stagnating, incomes seemed to be increasing (in the case of England as Broadberry et al. indicated). The only explanation is that workers increased their labor supply? Why would they do that? What happened that caused them to increase the amount of labor they were willing to supply? The arrival of new goods (sugar, tobacco etc.) caused them to change their willingness to work. This is a strong illustration of how preferences can change more or less rapidly (when new opportunities are unveiled). In fact, Mark Koyama (who blogs here) managed to insert this narrative inside a very simple restatement of Gary Becker’s model of time use. Either you have leisure that is cheap but time-consuming (think of leisure in the late middle ages) or leisure that is more expensive but does not consume too much time (think the consumption of tea, sugar and tobacco). Imagine you only have the time-expensive leisure which you value at level X. Now, imagine that the sugar and tea arrive and, although you pay a higher price, it provides more utility than the level and it takes less time. In such a context, you will likely change your preferences between leisure and work. I am grossly oversimplifying Mark’s point here, but the idea is that the industrious revolution argument advanced by de Vries can easily fit inside a simple neoclassical outlook. On top of solving many puzzles, it also shows that one does not need to engage in some fanciful flight of Marxian theory (I prefer Marxian to Marxist because it is one typo away from being Martian which would adequately summarize my view of Marxism as a social theory). If it fits inside the simpler model, then you don’t need the rest.  De Vries does just that.

  • Anderson, Terry Lee, and Peter Jensen Hill. The not so wild, wild west: Property rights on the frontier. Stanford University Press, 2004.

Governance is not the same as government (in fact, they can be mutually exclusive). In recent years, I have been heavily influenced by Elinor Ostrom’s work on how communities govern the commons in very subtle (but elaborate) ways without the use of coercion. These institutional arrangements are hard to simplify into one variable for a regression, but they are theoretically simple to explain: people respond to incentives. Ostrom’s entire work shows that people on the front line of problems generally have the best incentives to get the right solution because they have skin in the game. What her work shows is that individuals govern themselves (see also Mike Munger’s Choosing in Groups) by generating micro-institutions that allow exchanges to continue. Terry Anderson and Peter Hill provide the best illustration in economic history in that regard by studying the frontier of the American west. Settlers moved to the American West faster than the reach of government and the frontier was thus an area more or less void of government action. So, how did people police themselves? Was it the wild west? No, it was not. Private security firms provided most of the policing, mining clubs established property rights without the need for government, farmers established constitutions in voluntary associations that they formed and many “public goods” were provided privately. The point of Anderson and Hill is that governance did exist on the frontier in a way that demonstrates the ability of voluntary actions (as opposed to coercive government actions) to generate sustainable and efficient solutions. The book has a rich theoretical framework on top of a substantial body of evidence regarding the emergence of institutions. Any good economic historian should own and read this book.

  • Vedder, Richard K., and Lowell E. Gallaway. Out of work: unemployment and government in twentieth-century America. NYU Press and Independent Institute, 1997.

The last book on the list is an underground classic for me. Richard Vedder and Lowell Gallaway are very good economic historians. It was produced like many other underappreciated classics (like Higgs’s Crisis and Leviathan) by the Independent Institute (see their great book list here). Most of their output was produced from the 1960s to the 1980s. However, as the 1990s came, they moved towards the Austrian school of Economics. With them, they brought a strong econometric knowledge – a rarity among Austrian scholars. They attempted one of the first (well-regarded) econometric studies that relied on Austrian theory of the labor-market (a mixture of New Classical Theory with Austrian Theory). Their goal was to explain variations in unemployment in the United States by variations in “adjusted real wages” (i.e. unit labor costs) all else being equal. At the time of the publication, they used very advanced econometric techniques. The book was well received and even caught the attention of Brad DeLong who disagreed with it and debated Vedder and Gallaway in the pages of Critical Review. Although there are pieces that I disagree with, the book has mostly withstood the test of time. The core insights of Out of Work regarding the Great Depression (and many of its horrible policies like the National Industrial Recovery Act) have been conserved by many like Scott Sumner in his Midas Paradox and they feature prominently in the works of scholars like Lee Ohanian, Harold Cole, Jason Taylor, Price Fishback, Albrecht Ristchl and others. In the foreword to the book, they mention that D.N. McCloskey (then the editor of the Journal of Economic History) had pushed hard for them to publish their work regarding the 1920s and 1930s. McCloskey was right to do so as many of their contentions are now accepted as a legitimate (if still debated) viewpoint. The insights regarding the “Great Depression of 1946” (a pun to ridicule the idea that the postwar reduction in government expenditures led to a massive reduction in incomes) have been generally conserved by Robert Higgs in his Journal of Economic History article I mentioned yesterday (and in this article as well) and even by Alexander Field in his Great Leap Forward However, Out of Work remains an underground classic that is filled with substantial pieces of information and data that remains unused. There are numerous unexploited insights (some of which Vedder and Gallaway have followed on) as well. The book should be mandatory reading for any economic historian.

Minimum wage and length of poverty spells

I had a pre-programmed blog post on the issue of the minimum wage and poverty which was preempted by Mark Koyama (a blogger here at Notes on Liberty). The tweet is below and it has forced me to adjust the post.

Mark is absolutely right! Let me explain why with my own spin on it.

First of all, the demand curve slopes downwards – always. However, the method of adjusting to price changes (wages are a price and the minimum wage is a price control) is not an empirical constant.  I am unlikely to fire workers for a 1% in the inflation-adjusted minimum wage. Firing workers implies transaction costs that are dependent of context (for example, if I am friend with my employee, this is a transaction cost in the form of a lost friendship), firm size (I won’t fire my only employee which represents 50% of my output for a 1% hike in MW) laws (firing and hiring regulations), institutions (social institutions, reputation, norms), my clientele (how elastic is their demand) and technological alternatives. For a 1% increase, I am likely to reduce work hours or cut marginal benefits (no free soup for you). For a 10% increase, I am more likely to consider the option of firing a worker or I may shift to a new technological set that reduces my demand for labor.  It may happen rapidly or take some time, but there will eventually be an adjustment.

In any case, the minimum wage will imply some losses with a deadweight loss. Only the method by which it materializes is debatable.  By definition, some people will be hurt and generally and even if supply is super-elastic (doubtful), some suppliers (workers)  will be ejected from the market (or the quantity of labor they can supply will be ulitmately reduced). Since the minimum wage generally tends to fall on unskilled workers, this must be correlated with workers close to the poverty line.

Ideally, we’d need a measure of the minimum wage to be compared with the “at-risk” population over a long period of time in order to encapsulate all the effects of the minimum wage. The perfect measure is the “length of poverty spell” variable which has been emerging progressively from the BLS. The problem is that it is not broken down by state. Fortunately, Canada has that variable (well, a low-income variable which is a relatie poverty measure) for provinces. Inside the Survey of Labor and Income Dynamics (affectionally known as the SLID), this longitudinal variable has a span of eight years. Basically, we can know if a person has been below the low-income threshold for up to eight years. Let’s take that extreme measure and plot it against the minimum wage divided by the average wage.

As one can see from the scatter plot below, there is a more or less clear relationship between the minimum wage as a share of the average wage and the length of poverty spells. What is more impressive is that this graph is not a regression. More precisely, the provinces with the highest minimum wages (like my own province of Quebec and the province of Nova Scotia) also have the most extensive social welfare nets. Alberta,  a province with the lowest minimum wage ratio and one of the least “generous” social welfare net in Canada, is at the very bottom of the pack in terms of the persistence of poverty.

minimumwagepoverty

I think this graph acts in very modest (but clear) support to Mark’s point (which is also the point of Burkhauser, Sabia, MaCurdy and many others)

When Black Unemployment Rates Were Equal to White Unemployment Rates…

In a twitter-debate with Tariq Nasheed, I pointed out that the wages rates did converge between the 1940s and 1990s. Recently, Robert Margo of the University of Chicago extended this to per capita incomes since 1870. It is fascinating to see that there was convergence between 1870 and 1940 in spite of Jim Crow laws (it tells you how much more blacks could have achieved had the laws not existed – see notably the work of Bob Higgs on this).

income-convergece

Each time I see this evidence, I am bemused. You see, I often debate colleagues on particular features of social policy in order to assess policy reforms or the effects of past reforms. But, its always good to take a step back and look at the long-view of history. It puts things in perspective. The Margo graph does just that. It tells me the story of what could have been. And just for the sake of remembering properly (infer whatever conclusions you like), it is worth showing racial differences in unemployment rates since 1890. What strikes me is how similar the rates are until the 1950s. What happened at that point? When you ask yourself this question, you’re forced to put everything in perspective. And it becomes harder to have “generic” answers in the lazy-form of “its racism”. Why would racism explain the difference after 1950, but not before?
whites

Maybe, just maybe, people like Tariq Nasheed should stop proving that H.L. Mencken was right in saying that “for every complex problem, there is an answer that is simple, clear and plainly wrong”.

 

From the Comments: Why do France’s banlieues have 40% unemployment?

Dr Amburgey asks the question, and Dr J gives the answer:

Terry: Good question but the answer is implied: Policies that allow for much higher economic growth than has been the case since about 1985.

It’s hard to figure an explanation for persistent French economic stagnation that does not implicate government action (ACTION, not inaction). Two examples: Retail stores can only hold sales for twice two weeks in a year. (That’s as in “on sale.”) The government decides when the sales seasons take place all over France, at the same time, irrespective of local conditions. Yes, you read that right. Second example. An ideological battle has been running for at least ten years at the highest level of government about whether or not to allow large stores to be opened on Sundays. The pros just lost again! [but see Dr J’s update – bc]

I am a weak man, I can’t resist adding a third example: On Sunday mornings, you can buy delicious croissants in bakeries everywhere but they are not allowed to sell coffee! The cafes open late on Sundays. Dunking is effectively illegal in France for several hours every week.

The French political elite, almost all statist, seldom loses an opportunity to prevent employment from growing. Note that I am not especially blaming the current Socialist administration. There are almost no conservative parties in France today, have not been for many years. The word “libertarian” has no current French equivalent. (The French word “libertaire” is related but it means something else.)

French schools are mostly very bad. They are run by a centralized government bureaucracy.

Of course, economic stagnation is not about the children and grandchildren of immigrants specifically. It’s just that those least favorably positioned with respect to the job market tend to suffer most from stagnation. Children and grandchildren of immigrants are among those. If the French economy grew at an annual rate of say 2.5% – the current US rate, I think – even the children of immigrants in remote banlieues would see their employment opportunities multiply. At least, they could compete for something. There is not much leftist municipalities largely in charge of those immigrant-heavy areas can do, really. The best among them set up good soccer clubs, that’s about it.

Poor economic performance does not strike everyone equally. The offspring of immigrants are disadvantaged mostly for reasons that would not matter elsewhere, in Germany, next door, for example. I think racism and xenophobia play a small role. It seems to me that both were much much worse in the sixties and seventies yet, immigrants and their children had work then when the country’ s economy was growing at a normal pace.

Stagnation does not hit everyone equally: The outflow of graduates from the best schools (mostly engineering schools) is perceived to be so great that last year, the Socialist government created a new cabinet post for them. I suspect it’s to hold them back or to try and lure them back. Would I make this up?

Being an immigrant is just a potential basis for social organization (a la Marx). Being an immigrant from already secular Portugal or from Romania is not a good basis for such. Being an immigrant from a Muslim country (probably most immigrants to France) creates clear delineation because so much of Muslim culture is violated every day by ordinary French behavior. (Yes, some stereotypes are factually correct!)

Going back to your question about libertarianism specifically: I think that if 10 % of all government economic regulations were abolished suddenly, on a lottery basis, the French GDP growth rate would double immediately, with positive consequences for immigrants’ progeny, of course.

Terry you should read Delacroix (recently in Liberty Unbound). [“Religious Bric-à-Brac and Tolerance of Violent Jihad” – bc, again]

The rest of the thread is well worth reading, too, as Jacques and Terry size up each others’ views on the European Union.

Unemployment is Completely Unnecessary

In U.S. government statistics, a person is unemployed if he is 16 years of age or more, and that person is able and willing to work at prevailing wages. The labor force includes the employed and the unemployed. If one is not employed for wages because one does not wish to work or to seek work, that person is not in the labor force, and not counted as unemployed.

The unemployment rate in the USA is now about six percent, down from a peak of ten percent in 2009. About one percent of the labor force is in “frictional” unemployment, meaning that the worker is between jobs or recently graduated from school and engaged in job search, or about to be hired. When the economy is depressed, there is “cyclical” unemployment, those not working as firms reduce employment. There is also the “structural” unemployment of workers losing their jobs in declining industries, and the seasonal unemployment of those employed only during a season such as in resorts or during harvests.

An economy is in full employment when the only unemployment is frictional. The economic puzzle is why there is any other unemployment. Cyclical unemployment is no mystery, as firms have fewer sales as demand falls, and falling demands become a downward spiral as falling purchases by some become falling production by others. The recession ends when materials prices and real estate rentals have fallen so low that production becomes profitable again.

Since recessions are caused by monetary and fiscal subsidies, a pure market economy would have neither, so it would have no recessions and no cyclical unemployment. So the puzzle consists of chronic unemployment, those unable to obtain work even during prosperous times. Most of the unemployed have been out of work for months or years. Those long unemployed have even more difficulty finding employment, because employers wonder why that person can’t find any job.

Some economists consider idle labor to have a positive side. You car is not wasted when you don’t use it, because it provides the service of availability. Empty seats in a theater have value because the theater needs that capacity for popular shows. Likewise, in this viewpoint, idle labor provides workers when firms need to hire. Also, the unemployed need time to engage in job search, so they are busy even if unemployed. But one can be employed at least part time while looking for better work, and while idle labor may be good for employers, it is bad for workers who need the income, and for taxpayers who have to support those not working.

In a pure market economy, there would not be any unemployment at all. There would be no seasonal unemployment, because workers could find other jobs in other seasons. There would be no structural unemployment, because workers could shift to other industries, and work in temporary jobs while searching for full-time employment. Even workers in frictional unemployment would be able to work some of the time, since job search is not full-time.

One of the premises of economics is that human desires are unlimited. There is always a demand for something. That demand provides an opportunity for workers to be employed to satisfy that desire. In a pure market economy, one could also be easily self-employed. Any person who is not totally incapacitated would be able to offer some service at some wage. If the wage one can obtain is too low to bother with working, then that person would not wish to work, not be in the labor force, and not be unemployed.

Unemployment exists because there are barriers that prevent labor from having access to land and capital goods. If the cost of hiring a worker is greater than his productivity, he will not be hired. In a pure market, the wage would be set where the quantity of labor supplied by workers equals the quantity demanded by employers.

Government policy raises the cost of labor above the pure market wage. Minimum wage laws prevent employers from hiring the least productive workers. On top of the minimum wage are imposed costs: the employer’s share of payroll taxes, mandated medical insurance, worker accident insurance, and the unemployment compensation tax. The firm also has to withhold taxes from wages and send then to the government. There is also a litigation risk and cost of hiring labor, as labor laws promote excessive litigation to combat malpractice, discrimination and sexual harassment. Also, union labor monopolies, and laws favorable to unions, push up the wages of union workers at the expense of less employment. Finally, laws making it costly to fire workers raise the cost of hiring them, creating more unemployment.

In a full-employment economy, when firms seek to expand, they would pull workers away from other firms, or pulled into the labor force, by offering higher wages and better conditions. There is no need for idle labor.

The best policy for labor is full employment. Labor laws that seek to protect workers end up imposing barriers that prevent employment. Full employment requires hiring flexibility and the removal of government-imposed costs. Full employment requires the elimination of taxes on labor, exchange, production, and consumption. Public revenue from land rent or land value could replace all these labor-hampering taxes, while promoting the productive use of land which would further increase wages.

A shift in taxation from labor to land would both increase employment and increase wages, while letting the worker keep his wage. It is not unemployment that is a puzzle, but rather why workers are not demanding the abolition of their wage-tax burden.

The Theory of the Non-Working Class

In the USA, people of age 16 and above are considered of working age. Of those of that age range, those who are working, seeking work, or hired but not yet working, are designated to be in the labor force. The labor force participation rate is the number of people in the labor force divided by the number of those of working age.

From 1950 to 2000, the labor force participation rate in the USA rose from 59 percent to 67 percent. Much of that increase came from the doubling of the participation rate of women, from 30 percent in 1950 to 60 percent in 2000. But total labor participation has declined since 2000 to 63 percent.

While the portion of women in the US labor force rose, the portion of men has been declining. The prime working years are considered to be from age 25 to 54, and one sixth of the men of that age range are not working. In 1950, only four percent of men of that range were not employed.

Many of those not working are not seeking work, and are therefore not counted in the labor force. They are also not counted as unemployed, because by definition, the unemployed are those actively seeking work plus those who have been hired but not yet started to work for wages. Two thirds of working age men are not seeking work, although some who sought work but stopped because they were discouraged, would take a job if offered.

About 40 percent of the men seeking work have been unemployed for six months or more. The chronically unemployed are less likely to become employed, so the long-term unemployment feeds on itself.

The real wage of lower-skilled workers has been falling since 1970. For workers who did not finish high school, the real wage (adjusted for inflation) has fallen 25 percent. That fall in wages is offset somewhat by the availability of new products such as cell phones and by the fall in the relative prices of electronics and other goods, but the cost of housing, medical care, taxes, and college tuition have risen to offset some of that productivity gain.

There are several reasons why male labor participation has fallen. First, more men are attending college. Second, due to the expansion of the war on drugs, the portion of men in prison has risen. Third, as more women work for wages, some male partners choose home production, doing house work and child care at home, which is real labor but not counted in the output data. Fourth, more people are obtaining government’s disability income. Very few on disability go back to work. Fifth, many in the first of the baby-boom generation, born during 1946-1950, are retiring.

The downward trend of labor participation will continue. The Congressional Budget Office estimates that the participation rate will fall to 61% by 2024. CBO calculates that the Affordable Care Act reduce the labor force by more than 2 million jobs. Workers will be able to quit their jobs without losing medical coverage, and the expansion of Medicaid will induce many more adults to obtain medical care without having a job.

One of the problems with a lower labor participation rate is that it reduces the ratio of workers to non-workers. Social Security and Medicare are supported by transferring income from workers to non-workers. A smaller labor participation rate will use up the trust funds and create a deficit for these programs sooner. Also, fewer workers results in lower economic growth, which implies that more of those in poverty will stay that way.

Much of the labor participation decline is not voluntary, but caused by tax and subsidy policies. Without taxes on wages and enterprise profits, both wages and employment would be higher. If the funds now going into Social Security instead went into tax-free private retirement accounts, those who retire would rely on their own past savings rather than transfers from those working. Without the income-tax distortion caused by tax-free medical insurance and taxed money wages, workers would be able to choose the insurance plan that fits them best rather than having to accept the limited plans offered by employers and the government.

The best alternative to taxing wages is to tax land rent or land value. But even without such a fundamental shift in policy, the labor force participation rate can be made more voluntary with employee and self-employment incentives for those long out of work, such as tax offsets and exemptions from restrictions (e.g. licensing, union rules, and city zoning) that prevents working at home, and exemptions from litigation risks. Immigration reform – legalizing those already in the country and allowing more of those with labor skills into the country, would also substantially increase the labor population.

The basic problem with labor world-wide are restrictions on hiring and firing labor, and the heavy costs imposed by taxes, regulations, and mandates on employers. If an employer, including a self-employer, could simply hire a worker without having to deal with forms and regulations, and with no taxes on the employer and the employee, we would have full employment at wages that would provide a decent standard of living. The labor problems we have are iatrogenic, a disease caused by the doctor, in this case, the economic malady caused by government policy. The government people look to for solving economic problems has caused them in the first place.