Debate on workweek reduction

See also: Productivity and the Workweek

Posted by: Erik Rauch  Recipient: All
Posted: Wed, Mar 05, 2003 at 15:44:15 (EST)

Edgar Ayala wrote:
> Reducing the workweek doesn't really reduce unemploymet...
> it is more like taking unemployment and spreading it around.

In a manner of speaking, yes. But by unemployment I mean 'the number of people out of work' rather than the total number of hours worked. The justification is that it is better to have 4 people employed at 30 hours a week (at reduced incomes) than it is to have 3 people at 40 hours and 1 person at 0 hours (who must be supported by the state). It is true that such a redistribution involves some inefficiency (more training is necessary, for example) which leads to a one-time hit in productivity. This hit is probably not that large, and the partisans gloss over this, but the point is that reduced unemployment is worth this productivity hit. And in any case, the general march upward of productivity nullifies this one-time hit within a few years, even if it is not compensated by the reduction in unemployment benefits that must be paid.

> Taking me out of my job five hours early means there will be
> five hours where I would have been willing to put my labor
> towards the creation of goods and services (and be paid an
> income for it) but would not be able to.

Right.. but assuming that demand is constant, the business that is no longer allowed to hire you for those extra five hours will have to hire more workers, thus reducing unemployment (as defined above). Thus there is now someone working who, under the previous state of affairs, would have been willing to put his labor towards the creation of goods and services, but was not able to.

Now.. can someone tell me why workweek reduction proposals do not go over well with [economists such as Paul] Krugman and his colleagues? The only thing I can think of is that this attempted redistribution would lead to a reduction in demand for some reason. But I actually haven't seen any valid arguments for this yet. The 'lump of labor fallacy' doesn't imply that the limited workweek would reduce the total number of hours worked.
 
 

Posted by: Edgar  Recipient: Erik Rauch
Email Address: edgara @ mailmail.com

> Now.. can someone tell me why workweek reduction proposals do not go over well > with Krugman and his colleagues? The only thing I can think of is that this
> attempted redistribution would lead to a reduction in demand for some reason.

well it would lead to a reduction in demand because I am employing the most productive workers I can find. If you constrain their labor, I will be forced to hire those who are unable to find work otherwise. Thus, I will hire fewer workers.

> But I actually haven't seen any valid arguments for this yet. The 'lump of
> labor fallacy' doesn't imply that the limited workweek would reduce the
> total number of hours worked.

If by valid you mean arguments you don't like then well, are you right. I hate to sound like an economist for a minute here, but there is an obvious efficiency argument. The market has a solution for labor and you are constraining it. I must hire workers I don't want to hire because you won't let me hire Mr. Goodworker even though he would like to work some extra time.
 

Posted: Thurs, Mar 06, 2003 at 11:51:08 (EST)

> well it would lead to a reduction in demand because I
> am employing the most productive workers I can find.
> If you constrain their labor, I will be forced to hire those
> who are unable to find work otherwise. Thus, I will hire
> fewer workers.

Not necessarily. You could hire more workers to make up for the productivity hit.
 

> there is an obvious efficiency argument. The market has
> a solution for labor and you are constraining it. I must hire
> workers I don't want to hire because you won't let me hire
> Mr. Goodworker even though he would like to work some
> extra time.

I don't dispute that. I am saying that the market solution is socially undesirable, and that we can afford the productivity hit needed to fix it.
 

Posted: Thurs, Mar 06, 2003 at 18:56:01 (EST)
Original: Thurs, Mar 06, 2003 at 11:51:08 (EST)
Posted by: Edgar  Recipient: Erik Rauch
Email Address: eayala @ booyah.com

> Not necessarily. You could hire more workers to make up for the productivity hit.

You wouldn't. There is heterogeneity among workers. You first hire the most productive. When they reach their maximum hours, you will be hiring less productive workers. On average, those guys who are sitting around unemployed have an inferior skill set to those who are working. Rocket science it is not.
 

> I don't dispute that. I am saying that the market solution is socially
> undesirable, and that we can afford the productivity hit needed to fix it.

If you don't dispute that it is a less efficient outcome, then why go around saying that economists have no justifciation for not doing it?
 

Posted: Fri, Mar 07, 2003 at 00:12:14 (EST)
Original: Thurs, Mar 06, 2003 at 18:56:01 (EST)
Posted by: Erik Rauch  Recipient: Edgar
Email Address: erauch @ ai.mit.edu

> You first hire the most productive. When they reach their
> maximum hours, you will be hiring less productive workers.

Yes, but this still doesn't say why demand would significantly decrease. If demand were constant you would actually have to hire more of them to produce the same.

> If you don't dispute that it is a less efficient outcome, then
> why go around saying that economists have no justifciation
> for not doing it?

Why? Because GDP is not everything. An economy which is more efficient at producing goods and services with monetary value is not the same as an economy that provides a better living.

Surely this distinction is accepted to some degree by most economists, as I don't see many of them calling for repeal of time-and-a-half for overtime, for example. However, this distinction seems to be forgotten even by some of our most perceptive economists, including Krugman sometimes. The discipline of economics cannot be merely the science of increasing GDP growth, as this can impact people negatively as well as positively, and a decrease in efficiency may be a social gain.
 

Posted: Fri, Mar 07, 2003 at 07:37:19 (EST)
Original: Fri, Mar 07, 2003 at 00:12:14 (EST)
Posted by: edgar ayala  Recipient: Erik Rauch
Email Address: edgara @ booya.com

> Why? Because GDP is not everything. An economy which is more efficient at
> producing goods and services with monetary value is not the same as an
> economy that provides a better living.

You seem to be trying to present a classic equity v. efficiency argument...or taking a normative position and presenting it as if it were positive. This is not the same as economists having no reason to oppose shorter workweek laws. Economists do have reasons. Good reasons. You just don't like them. It is one thing to say 'I think economists tend to overemphasize efficiency relative to equity,' and another to say, 'Economists have no reason to oppose shorter workweek laws' Yours is a value judgement.

> However, this distinction seems to be forgotten even by some of our
> most perceptive economists, including Krugman sometimes.

Not forgotten. They just apparently value efficiency relative to equity more than you wish they would.

> The discipline of economics cannot be merely the science of increasing GDP
> growth, as this can impact people negatively as well as positively, and a decrease
> in efficiency may be a social gain.

I don't know where you are getting this strawman 'discipline of economics.' See, there is this dude named Amartya Sen....
 

Posted: Fri, Mar 07, 2003 at 18:15:15 (EST)
Original: Fri, Mar 07, 2003 at 07:37:19 (EST)
Posted by: Erik Rauch  Recipient: edgar ayala
Email Address: erauch @ ai.mit.edu

> It is one thing to say 'I think economists tend to overemphasize
> efficiency relative to equity,' and another to say, 'Economists
> have no reason to oppose shorter workweek laws' Yours is a
> value judgement.

But so is theirs. There are cases when GDP growth can come at the expense of social criteria, and so their advocacy in these cases is laden with value judgements. Economics cannot be free of them. When someone like Krugman argues for a policy, he doesn't just imply that it is just because it would increase growth, he implies it would be a better policy, period.
 

Posted: Fri, Mar 07, 2003 at 07:20:32 (EST)
Original: Fri, Mar 07, 2003 at 00:12:14 (EST)
Posted by: Edgar  Recipient: Erik Rauch
Email Address: edgara @ booyah.com

>> You first hire the most productive. When they reach their
>> maximum hours, you will be hiring less productive workers.

>Yes, but this still doesn't say why demand would significantly decrease. If demand
> were constant you would actually have to hire more of them to produce the same.

Because if you hire workers until their marginal productivity of labor no longer exceeds the prevailing wage, dealing with a pool of workers with lower marginal productivity means we meet that threshold earlier.
 

Posted: Wed, Mar 05, 2003 at 17:25:48 (EST)
Original: Wed, Mar 05, 2003 at 15:44:15 (EST)
Posted by: Bobby  Recipient: Erik Rauch
Email Address: robert @ pkarchive.org

Supply increases probably do increase demand if you control for everything else but in reality are swamped by other things. Krugman also said in Accidental Theorist that it was a monetary phenomenon is demand is too small, and he assigns that job to the fed. I'm not so sure economists criticize the French plan for failing to stem unemployment. In fact it probably does do just that if you control for other variables. However there are three much better solutions to the unemployment problem (1) fiscal stimulus (2) monetary stimulus (3) and market reforms that make hiring and firing less costly to employers. The French plan is an unnecessary productivity hit, while better answers exist. There is no reason understate the severity or duration of this plan's effects on producivity growth. Granted it's effects on growth would taper off eventually, but when? Next year? next decade? Next century? Next millenium? I don't know.
 

Posted: Thurs, Mar 06, 2003 at 12:22:13 (EST)
Original: Wed, Mar 05, 2003 at 17:25:48 (EST)
Posted by: Erik Rauch  Recipient: Bobby
Email Address: erauch @ ai.mit.edu

> However there are three much better solutions to the
> unemployment problem (1) fiscal stimulus (2) monetary
> stimulus (3) and market reforms that make hiring and firing
> less costly to employers. The French plan is an
> unnecessary productivity hit, while better answers exist.

It seems like you would want to use any method at your disposal. Those 3 things are unlikely to solve the whole problem (and many of the French may consider the security of their regulations related to hiring and firing to be worth it). That's why I don't think you could call it 'unnecessary'.

> There is no reason understate the severity or duration of
> this plan's effects on producivity growth. Granted it's effects
> on growth would taper off eventually, but when?

Here is the heart of my argument: productivity growth is mostly technological, so any hit caused by redistributing work hours among the population would be a one-time hit. That is, it would cause us to become less productive than we otherwise would have been (as businesses spend more on training and hire less productive workers), but not continually cause us to become ever less productive over time. So how big would the reduction be - 2%? 5%? 20%? In any case, consider that productivity grows at between 1% and 3% per year. Within a few years, this will have totally nullified the reduction in productivity. Thus, you would trade just a few years' worth of potential gain in incomes for the extra employment and reduced hours. Did people in 1997 (15% lower productivity according to BLS) or even 1990 (25% lower) really have it so bad?
 

Posted: Thurs, Mar 06, 2003 at 14:54:37 (EST)
Original: Thurs, Mar 06, 2003 at 12:22:13 (EST)
Posted by: Bobby  Recipient: Erik Rauch
Email Address: robert @ pkarchive.org

Actually Krugman guesses that 1 or two points could be knocked off of France's unemployment if the Maastrict treaty did not require such fiscal restraint and a strong franc (this of course has to be updated for the new EuroZone, where deflation is a possible problem). Without these labor market rigidity problems that I described, most economists think that yes this unemployment problem WOULD be solved. I unfortunately do not know any of the models describing this so I can't elaborate. Also for all we know, the rigidity that this new regulation imposes on the labor market could increase unemployment. I don't know if this is in the plan, but if the government makes employers pay the same salary for less hours, this would create a major excess labor supply problem that would exacerbate the unempoyment problem similar to a drastic minimum wage. If this were so, the law would accomplish nothing in unemployment as well as hurting productivity. Likewise it would price low-skilled workers out of the labor market and change the composition of employment. Also, the French public like the American public does not adequately understand economics, and we shouldn't judge how they would judge their own regulations if they knew the trade-offs involved. French politicians have always complained about the severity of their employment problem, which Krugman says is by far worse than any of their other economic problems, but blame it on irrelevant things like currency speculation and globalization. You can read about it in this article: http://www.pkarchive.org/global/gauls.html

World technological progess is necessary but not a sufficient for a country's productivity growth. It must also be as easy as possible for a country's firms to employ those new technologies. This new regulation might make it so that France might not be able to employ some percentage of the new technologies developed each year, which WOULD hurt its growth rate. By the way, the effect on growth would probably not show up as one big one-time thing but just a smaller growth rate for many years. With this policy it could take decades for France to get back to the same growth at which world technological progress grows that is if it actually can do that. Would the French be willing to put up with 10, 20, 50 or 100 years of sub par growth relative to the rest of the world a result of this policy (that is if the public had an understanding of economics)? Another problem is that this policy would decrease France's productivity level and income relative to other countries permanently, which has the problem of (A) national pride and (B) its government is not allowing France's economy to do as well as it can. Exponential growth also implies ever increasing income differentials between countries with same growth.

In short this policy's ability to help the unemployment problem is at best dubious, whereas all of its other consequences especially on productivity are bad. As policies go, you could do much better namely the three I mentioned before, any one of which would avoid this productivity problem and definitely address the problem at hand which is unemployment. Income growth in the US has not been distributed evenly, so the median income growth is probably way less than the numbers you cited.
 

Posted: Thurs, Mar 06, 2003 at 23:46:41 (EST)
Original: Thurs, Mar 06, 2003 at 14:54:37 (EST)
Posted by: Erik Rauch  Recipient: Bobby
Email Address: erauch @ ai.mit.edu

> Also for all we know, the rigidity that this new regulation
> imposes on the labor market could increase unemployment.
> I don't know if this is in the plan, but if the government
> makes employers pay the same salary for less hours, this
> would create a major excess labor supply problem that
> would exacerbate the unempoyment problem

In the French example, employers aren't required to keep the total salary the same. I agree it might not work to promise reduced hours on the same salary. However, if productivity means anything, you can promise this: fewer hours for the same standard of living of a 1997 (or 1995 or 1990) worker who worked 40 hours.

> This new regulation might make it so that France might not be
> able to employ some percentage of the new technologies
> developed each year, which WOULD hurt its growth rate.

What would give you reason to believe that? I think there is reason to believe the contrary: with all the other labor market regulations Europe has, European productivity growth has not lagged behind that of the US, a fact Krugman has pointed out. In fact, measured over the past 12 or 25 years, productivity growth has actually been higher than in the US (see Robert Gordon's data).
 

> By the way, the effect on growth would probably not show up as
> one big one-time thing but just a smaller growth rate for many
> years.

Well, as long as that number of years is finite, it's the same as a one-time hit. I don't have a way to know how big it would be -- does anyone? But I think it's unreasonable to think it would be more than 25%, which is 12 years' worth of growth, and probably much less. With the French reduction of 39 to 35 hours, a maximum of 10% of all the hours worked are affected. Even if those hours were 50% less productive (because they are worked by less productive workers who need training), this would mean a 5% hit. Taking the workweek down to 30 hours would affect 23% of the hours, and lead to a 12% hit if the redistributed hours were half as productive.
 

Posted: Fri, Mar 07, 2003 at 12:18:11 (EST)
Original: Thurs, Mar 06, 2003 at 23:46:41 (EST)
Posted by: Bobby  Recipient: Erik Rauch
Email Address: robert @ pkarchive.org

What you seem to be arguing is that static changes aren't very important because they will be cured by the wonders of exponential growth. Economists know a lot about exponential growth, so why do they spend so much time on deregulation schemes, tax policy, free trade, etc. that create permanent increases in national income but do not increase its growth rate in the long run? Have economists failed to realize this insight? I think one reason they are concerned about static questions is that they want the economy to be on the greatest balanced growth path as possible for some target amount of unemployment, equity, etc. By this criterion, much better policies exist than this French policy, which is probably a fourteenth or fifteenth best solution to France's unemployment problem, especially compared to the three I discussed before. The French policy also fails the Pareto criterion. It is still not clear to me why you think it would decrease unemployment in the first place.

Why are mandated decreased work hours a good thing? Why shouldn't people be allowed to choose to work as much as they want for the wage that is offerred? The deadweight losses from this plan seem great to me.

By the way, the last time that the U.S. lost 25% of its GDP it was called the Great Depression (I don't know the per capita figure), and I'm not sure that France would be too happy with something like that even if it occurs in slow motion through decreased growth (see Japan for slow motion depressions). Deviations from trend in GDP on the supply side also cause huge unemployment problems (as well as inflation if the central bank does not intervene), and the French can do better. By looking at any recession you can see that losses of GDP do much more harm than the good that was achieved when that GDP was originally created.

'with all the other labor market regulations Europe has, European productivity growth has not lagged behind that of the US, a fact Krugman has pointed out. '

There is no reason to attribute Europe's productivity growth to labor market regulations. For all we know productivity could have been better without them. A more thorough examination is needed, and my sense without any proof is that economic theory would suggest this.
 

Posted: Fri, Mar 07, 2003 at 18:02:06 (EST)
Original: Fri, Mar 07, 2003 at 12:18:11 (EST)
Posted by: Erik Rauch  Recipient: Bobby
Email Address: erauch @ ai.mit.edu

> What you seem to be arguing is that static changes aren't very
> important because they will be cured by the wonders of
> exponential growth.

Well, what I'm saying is that productivity growth will make up for any reasonable efficiency loss that you incur by satisfying other criteria, as long as you don't reduce productivity growth too much. This is the same argument I make to computer scientists that I know. They have a choice between programming in a fast but inelegant language, or a slower but elegant language which is more humane to program in. I tell them they should program in the slow one, because just a few years of Moore's law will make it just as fast as the fast one is today. But they seem to want the absolute fastest thing available right that minute.

> Economists know a lot about exponential growth, so why do they
> spend so much time on deregulation schemes, tax policy, free
> trade, etc. that create permanent increases in national income
> but do not increase its growth rate in the long run?

I think it's because they're trying to reduce unemployment,. I'm arguing that they are missing solutions that would reduce unemployment at the cost of some efficiency, because they are fixated on GDP as a measure of goodness.

> The French policy also fails the Pareto criterion.

It depends what yardstick you use. Progressive taxation also fails the Pareto criterion if you simply count total income and not equity.

> Why are mandated decreased work hours a good thing? Why
> shouldn't people be allowed to choose to work as much as they
> want for the wage that is offerred?

Because in their economy, the market seems to select an outcome where 8-10% of workers are unemployed, and limiting the hours of some is a way to help alleviate that. It's the same spirit as progressive taxation - why everyone shouldn't be allowed to keep all of the money they earn - or the work redistribution laws that we already have (the 40 hour week).
 

> Deviations from trend in GDP on the supply side also cause
> huge unemployment problems .. By looking at any
> recession you can see that losses of GDP do much more
> harm than the good that was achieved when that GDP was
> originally created.

But all the data that inference is based on is from an economy with the current workweek, which hasn't been changed since the 30's. There is a correlation between GDP growth and unemployment, certainly, but the relationship is not necessarily causal, so this doesn't tell us what a shorter week would do.

> There is no reason to attribute Europe's productivity growth
> to labor market regulations. For all we know productivity could
> have been better without them.

I suppose, but this objection seems weak - Europe has both more regulations and faster long-term productivty growth than the US; that seems like good evidence that regulations don't necessarily reduce growth.
 

Posted: Fri, Mar 07, 2003 at 21:05:25 (EST)
Original: Fri, Mar 07, 2003 at 18:02:06 (EST)
Posted by: Bobby  Recipient: Erik Rauch
Email Address: robert @ pkarchive.org

You are seeing a false choice and making the tradeoff between unemployment and other things like GDP and efficiency more dire than it has to be. My main argument is that there are better policies with the same equity and unemployment associated with them, and WITHOUT these costs of efficiency, inconvenience, and restriction on choice associated with them.

Economists do not make policies that 'create permanent increases in national income but do not increase its growth rate in the long run' in order to fight unemployment. In fact the issue of increasing the economy's productive capacity usually has little to do with unemployment. Absent supply shocks that decrease the economy's productive capacity, in the short run unemployment is decided by the central bank with an error and in the long run the natural rate of unemployment is decided by issues of labor market flexibility. They focus on these static efficiency issues so that the pie is as big as possible and there is as much aggregate social welfare and national available as possible. Then equity concerns can be taken care of through post market redistributions.

Progressive taxation can do pretty well by the Pareto criterion in terms of redistributing producer and consumer surplus with small DWL. You can argue that progressive taxation PARETO DOMINATES a more regressive system because both compensated and uncompensated labor supply become steeper as the wage increases. therefore there might be less deadweight loss for the same amount of revenue gained in a progressive system compared to a regressive one.

It is not the free market that is responsible for French high structural unemployment. It is labor market rigidity due to their job security and other labor market regulations. This new French policy just trades in some unemployment for underemployment that is if it even does anything for unemployment.

But there are better ways to decrease unemployment without sacrificing equity and without sacrificing efficiency and without putting france on an inferior balanced growth path. If the French want to get lower unemployment, they should abolish some of these regulations. Any equity or income distribution problems can be taken care of an appropriate system of progressive income taxe and some combinatiion of wage subsidies akin the our Earned Income Tax Credit and direct income support. If resource misallocation across markets is a problem due to oversupply of labor wage subsidies can be corrected direct income transfers that would make them decrease their labor supply to the amount that would occur without any intervention. The result would be that aggregate welfare would be increased and appropriate redistributions could be made from winners to losers so that everyone on average is better off. Even without such redistributions, if the French understood that they could get less unemployment without decreased hours by abolishing some of these regulations I am sure that they would trade away some of them. I am not sure why we have been assuming that these labor market regulations are so sacred.

Unfortunately the public is ignorant of economics, and politicians like to blame unemployment on things like globablization that are irrelevant to the unemployment rate. The point is that this policy we've been discussing and many other French labor market regulations impose unnecessary costs upon society, whereas alternative policies exist that could achieve the same goals without such costs. My point is that WE CAN DO BETTER BY EVERY CRITERION: UNEMPLOYMENT EFFICIENCY GDP and all for the same equity. These words are indeed too extreme, but something like an approximation to this ideal can be made.

It can not be understated that the big problem is that we have ignored business cycle issues in this discussion, and those issues would probably take care of much of the contriversy over present unemployment. Right now isn't France in an economic slump facing deflation? If so, unemployment is worse than usual right now since inadequate demand is making unemployment greater than its natural rate. Proper monetary or fiscal operations would decrease unemployment back to the (irregularly large) natural rate where it usually is, unemployment would become less of a concern, and this reckless workweek proposal would not be taken seriously anymore.

I also think we should note that work hours have been and continue decreasing over this past century not due to regulations but mainly since people are consuming more leisure with growing incomes.

Krugman himself said that this policy proposal decreases France's productive capacity. I do not know, but this sounds like the temporary effect of this policy might look a lot like an oil shock, which does increase unemployment

My objection regarding productivity is not weak at all. Coincidence is not anywhere near enough proof for causation and should not be mistaken for such. As Paul Krugman said, 'good things happen to bad ideas.'
 

Posted: Sun, Mar 09, 2003 at 23:39:02 (EST)
Original: Fri, Mar 07, 2003 at 21:05:25 (EST)
Posted by: Erik Rauch  Recipient: Bobby
Email Address: erauch @ ai.mit.edu

> The point is that this policy we've been discussing and many other
> French labor market regulations impose unnecessary costs upon
> society, whereas alternative policies exist that could achieve the
> same goals without such costs. My point is that WE CAN DO
> BETTER BY EVERY CRITERION: UNEMPLOYMENT
> EFFICIENCY GDP and all for the same equity. These words are
> indeed too extreme, but something like an approximation to this
> ideal can be made.

Why do you assume that French labor market regulations are worthless? I would like you to consider that the costs may not be 'unnecessary'. Job security is a virtue not reflected in GDP statistics, and not much of it occurs without regulation. One may argue about whether the specific regulations in European countries accomplish this, but I am talking about the principle.

Unfortunately, it appears such regulations may contribute to unemployment, so you adopt counter-measures such as redistributing work hours. Yes, the economy is somewhat less productive that way but it is also richer in values not measured by the market. You may argue that job security is not worth the reduced national income, but that is a value judgement. And it is, I believe, an untenable one, because productivity keeps increasing regardless and renders your argument increasingly irrelevant.
 

Posted: Mon, Mar 10, 2003 at 15:55:37 (EST)
Original: Sun, Mar 09, 2003 at 23:39:02 (EST)
Posted by: Bobby  Recipient: Erik Rauch
Email Address: robert @ pkarchive.org

Job security is a good that is consumed by members of society just like low unemployment, high GDP, income equality, high static aggregrate welfare, etc. The question is what is the proper mix of all these goods, and a well informed citizen would usually want his society to have a diverse mixture of all of these rather than holding one of them sacrosanct. Moreover, I think that a properly designed social safety net can be a good substitute for job security, and France has a huge safety net (but is it properly designed?).

I also do not see why the job security situation would be so bad without these regulations. If you have two choices: Put several extra points on your natural rate of unemployment and make it impossible to change to another job (or to get a new one if by chance you lose yours) since they just aren't available but those who have jobs keep them permanently or (2) the average job last no longer than, say, 8 years but it is easy to find a new one. I would choose (2).

I am not sure why we should focus on only on regulations that decrease the probability that an unemployed person will be offerred a job, which increase the unemployment rate. From the emphasis of the textbooks, I get the sense that the generosity and design of unemployment insurance programs is probably a MUCH bigger issue in terms of increasing the natural rate of unemployment. The generous unemployment program makes unemployment an attractive option and increases the reservation wage at which workers would accept a job and makes it less likely that they would accept those scarce jobs. This is not an attack on the welfare state as a whole but really the incentives that it provides workers. This is a very big reason why I think that the work-sharing would not increase employment that much. If these jobs at these wages weren't attractive before work-sharing, why would they be attractive after? What if no one takes these jobs and this work-sharing proposal creates a labor shortage and inflation?

Concerning work-sharing, I also think that an argument could be made that too few hours is almost as much of a problem as too few people with jobs (especially since there is a cushy welfare state to soften the blow of no job).

If you are faithful to exponential growth, I also don't see why you should not just do nothing, let the economy grow without this disruption, and redistribute some of the national income to those who are unlucky enough to be employed. But that is already what is happening in France (and the rest is the business cycle). Don't make the employed, who have mortgages to pay, children to feed, tuition to pay, miserable for seven years or whatever by decreasing their hours and wait for per capita growth to cure everything.

I also think that years of forgone consumption as a result of this policy is a cost and is not unimportant. You also move ever farther away from what you could have achieved without the regulation as you grow exponentially.

If I've created the impression I am thinking about growth in terms of only measured GDP, I am not. Measured GDP can be used as a very rough proxy for growth of consumer and producer surplus, but is far from perfect and measures about half of what the ideal measure should capture. It is the expenential growth of CS and PS that I am concerned about, when I discuss static and dynamic inefficiency. Things like aggregate fairness and social values can be shown as public goods that the government corrects for and can be built into this CS and PS framework that I am considering. The idea is that the policy maker is ideally to get as much social surplus as possible and redistribute according to a fairness criterion, but the reality of policy making is obviously way more rough than this.

This is what I mean when I say:I think one reason they are concerned about static questions is that they want the economy to be on the greatest balanced growth path as possible for some target amount of unemployment, equity, etc

Remember that, from the viewpoint of conditional convergence, the U.S., Mexico, South Africa, and Afghanistan all grow at the same rate. The income differentials occur because each country is on a different balanced growth path from the other. I would never argue that we should not worry about Mexico or Latin America because 2% growth will eventually cure all.
 

Posted: Mon, Mar 10, 2003 at 19:54:39 (EST)
Original: Mon, Mar 10, 2003 at 16:57:54 (EST)
Posted by: Erik Rauch  Recipient: Bobby
Email Address: erauch @ ai.mit.edu

> Remember that, from the viewpoint of conditional convergence,
> the U.S., Mexico, South Africa, and Afghanistan all grow at the
> same rate on their different balanced growth paths. The income
> differentials occur because each country is on a different balanced
> growth path from the other.

Indeed, but here the difference is big enough to matter. My argument applies to rich countries, because they have long since passed a point of diminishing returns. The same increase in GDP does much less good now than it used to. I don't just mean because of increasing income equality; this applies to Europe also, where inequality hasn't risen nearly as fast. I mean the utility of additional increases in income is decreasing. This applies even if you measure percentages: increase the income of an Indian farmer by 10% and he can buy a water pump and won't have to carry the water; increase the income of a suburban American by 10% and he can buy an SUV instead of a mid-size car. The former is a vastly greater improvement. That's why I'm suggesting increases in productivity can be used in ways other than increasing income - there are things that would be a much better improvement.
 

Posted: Mon, Mar 10, 2003 at 19:52:43 (EST)
Original: Mon, Mar 10, 2003 at 15:55:37 (EST)
Posted by: Erik Rauch  Recipient: Bobby
Email Address: erauch @ ai.mit.edu

> If you have two choices: Put several extra points on your natural
> rate of unemployment and make it impossible to change to another
> job (or to get a new one if by chance you lose yours) since they just
> aren't available but those who have jobs keep them permanently or
> (2) the average job last no longer than, say, 8 years but it is easy to
> find a new one. I would choose (2).

But because both job security and ease of hiring and firing have value, I think the right balance is between your two extremes, and this is to some extent what they have in Germany and France (though I am not saying it is perfect).

> Job security is a good that is consumed by members of society
> just like low unemployment, high GDP, income equality, high
> static aggregrate welfare, etc. The question is what is the proper
> mix of all these goods, and a well informed citizen would usually
> want his society to have a diverse mixture of all of these rather
> than holding one of them sacrosanct.

Yes, that is what I am proposing. I believe that economists do just what you are criticizing: they pay too much attention to income, even when they do care about equality, and overlook values outside the market.

> The generous unemployment program makes unemployment an
> attractive option and increases the reservation wage at which workers
> would accept a job and makes it less likely that they would accept
> those scarce jobs. If these jobs at these wages weren't attractive before
> work-sharing, why would they be attractive after?

For the same reason that work-sharing may cause employers to have to hire less productive workers, it makes a wider and better choice of jobs available to the unemployed. It is two sides of the same thing.

> I also think that years of forgone consumption as a result of this policy
> is a cost and is not unimportant. You also move ever farther away from
> what you could have achieved without the regulation as you grow
> exponentially.

No, you don't, as long as productivity growth doesn't decrease. A small decrease would make the gap widen, but very slowly.
 

Posted: Mon, Mar 10, 2003 at 21:43:15 (EST)
Original: Mon, Mar 10, 2003 at 19:52:43 (EST)
Posted by: Bobby  Recipient: Erik Rauch
Email Address: robert @ pkarchive.org

My comment on forgone consumption was that the difference between the consumption or social surplus or whatever that you could achieve without the regulation and what you achieve with the regulation with it is persistent and grows exponentially. If you look at two economies that begin at different initial levels but they both grow at the same rate, the differential between them also grows at that rate. Which is to say that the forgone income, consumption, social surplus as a result of these regulations grows exponentially. Therefore static losses of efficiency are VERY important.

I do not understand your point about my comment being the other side of the coin. I was saying there is no point in artificially creating more jobs if people are not going to take them since the dole is a better option. And if the major cause of this abnormal natural rate of unemployment is this unemployment welfare problem as opposed to the rigidity problems, this work-sharing does nothing except create an artificial labor shortage.

Concerning you're other post below. By the same token, France has lots of equality and fairness, lots of unemployment, little static efficiency and flexibility thoughout its economy. In order fight this persistent natural rate of unemployment, why trade in that which you already have very little of when you could trade some of what you already have lots of? My point about growth was that small initial static differentials can grow over time into staggeringly huge ones.

8 years to have a job seems pretty reasonable for job security, so I would choose (2) and am not sure why you would trade any of the ease of hiring and firing away for more security not to mention the other hardships I mentioned.

This business about economists being obsessed too much with GDP is BS perhaps trumpeted by the media's idiotic reporting on the economy. You can find alternative gdp measures at William Nordhaus' homepage. You'll notice that newpapers always quote 'economists' who at a closer look don't have anything more than an MBA or in James Glassman's case a bachelor's degree. These people are not really economist, just talking heads that make noise. They get far more press than PhD economists, and I think this could be an expanation.
 

Posted: Tues, Mar 11, 2003 at 14:43:11 (EST)
Original: Mon, Mar 10, 2003 at 21:43:15 (EST)
Posted by: Erik Rauch  Recipient: Bobby
Email Address: erauch @ ai.mit.edu

You say that France has 'tons of inefficiency'. According to a paper from the US Bureau of Labor Statistics, France has significantly higher labor productivity than the US! So do Belgium, The Netherlands, and Norway. Despite its significantly lower working hours, France is about even with us per employed worker rather than per hour. If you count the unemployed, it's about 10% less.

Robert Gordon's data for Europe shows that the productivity gap between Europe and the US has been closed:

Posted: Tues, Mar 11, 2003 at 19:07:08 (EST)
Original: Tues, Mar 11, 2003 at 14:43:11 (EST)
Posted by: Bobby  Recipient: Erik Rauch
Email Address: robert @ pkarchive.org

I was talking about a lot of inefficiency as measured by DWL NOT GDP. It is possible to have a lot of GDP and DWL. Comparing French to American GDP is not the right comparison. You must compare what France could have achieved without the regs to France with the, by measuring the DWL. Perhaps international comparisons of DWL would be appropriate.
 

Posted: Tues, Mar 11, 2003 at 19:32:46 (EST)
Original: Tues, Mar 11, 2003 at 19:07:08 (EST)
Posted by: Erik Rauch  Recipient: Bobby
Email Address: erauch @ ai.mit.edu

But with 'deadweight loss' you are still dealing with static differences. Productivity growth can make up for it in a short time, unless the difference is extremely large.
 

Posted: Tues, Mar 11, 2003 at 20:04:45 (EST)
Original: Tues, Mar 11, 2003 at 19:32:46 (EST)
Posted by: Bobby  Recipient: Erik Rauch
Email Address: robert @ pkarchive.org

But the static DWL is a persistent cost and must have attention. Also if GDP is a good proxy for social surplus and as GDP grows social surplus grows. You can think of the persistent and growing shortfall in GDP from what it could have been without regulation as an ever growing shortfall in social surplus from what it could have been without the regulation. My point is that comparing French productivity GDP or social surplus to the U.S. does not sya very much. You must compare French GDP social surplus or productivity to what it could have been in France without the regulation. And please control for irrelevant variables.
 

Posted: Tues, Mar 11, 2003 at 14:41:50 (EST)
Original: Mon, Mar 10, 2003 at 21:43:15 (EST)
Posted by: Erik Rauch  Recipient: Bobby
Email Address: erauch @ ai.mit.edu

> The criticism that economists 'pay too much attention to
> income' is a stereotype that is not true and is currently being
> addressed in the profession with new measures of national
> income (see Bill Nordhaus' homepage for starters). I suspect
> that this stereotype comes from the media and that there are
> so many people on Wall Street who call themselves and are
> quoted in the news as economists but have nothing better
> than a bachelor's degree like James Glassman or MBA.

This is a straw man. You are attacking people like William Greider (a number of Krugman's criticisms of his book are on-target, I think), and not the specific arguments I gave. I am criticizing economists' unwillingness to consider the idea of trading some amount of static efficiency for other benefits, especially those that are real but hard to model.

New measures of national income are a good thing, but economists also must be concerned with things that are hard to measure and model.

> My point about forgone consumption is that this policy lowers
> the initial value from which the French economy grows, which
> involves very much forgone consumption. If you look at two
> economies on growth paths that increase at the same growth
> rate but have different initial values the vertical distance
> between them increases.

Now I think I see where you are getting this idea that the difference gets worse over time. While what you say about the 'vertical distance' is true, it doesn't matter. What matters is the horizontal distance. Let's suppose that a country whose productivity grows at 2.5% implements a 30-hour week and efficiency decreases by 15% as a result (let's suppose that it's immediate for the sake of example). This is 6 years' worth of growth, so their productivity has been knocked back to its 1997 level. If growth continues for another six years, they will be exactly back where they started. They gave up six years worth of growth, and will always be 6 years behind where they otherwise would have been, or behind a similar country which didn't make the change. This difference doesn't grow over time.

This foregone increase is just not worth worrying about considering the benefits.

> Who cares if the jobs are available if living off the dole provides
> incentives for people not to take them. I did not understand your
> idea about the other side of this coin . . .

You are saying that people may prefer to collect unemployment benefits rather than work at jobs that are available. First of all, these benefits may be comparatively generous in much of Europe, but they are still low compared with the average paying job. Most people would prefer to work unless the jobs available are very undesirable. This may be due to money, but may also be that they don't like the available jobs. (That is assuming they are available at all; in some cases even a motivated person cannot find one.) But my argument is this: reducing the workweek forces existing jobs to be redistributed to a greater number of people. Even if someone were unemployed before because there were jobs available but they didn't like them, there are now many more chances that there is a job that they will find preferable to unemployment benefits, so unemployment will decrease.
 

Posted: Tues, Mar 11, 2003 at 19:55:14 (EST)
Original: Tues, Mar 11, 2003 at 14:41:50 (EST)
Posted by: Bobby  Recipient: Erik Rauch
Email Address: robert @ pkarchive.org

I did address your criticism. It's just flat wrong. It's not true among academic economists. They are concerned with that which is hard to model, and they find a way to model it or to discuss it intelligently. Do you know who Amartya Sen is?

People don't gain pleasure from growth itself so 6 years of forgone growth is NOT the cost. It is consumption that beings brings pleasure. My point was that yearly forgone consumption is the cost of these policies. And that cost grows exponentially and that is a huge cost to any policy that creates DWL. You are not paying attention to this HUGE opportunity cost of these ideas, and you did not disprove that point. Who cares if your income now is always the same as what it could have been six years ago? It does not change the fact that the opportunity to consume in those past years and in the present one is lost. That is the cost of the plan. The good that is forgone is consumption.

You are in effect arguing that this work-sharing decreases the reservation wage of those on unemployment insurance since the array of jobs is wider seems like wishful thinking to me. To think that this would be a large effect seems to me like very wishful thinking and it would probably be swamped by other effects of this policy that would be just as large. For example, if I'm an employer who must hire in ways that I find less attractive, maybe I'll substitute away from labor to other inputs and employment will decrease. So I think that this unemployment insurance would be a HUGE obstacle to this policy having any effect on unemployment.

There are theoretically and empirically sound ways of decreasing the natural rate of unemployment such as reforming unemployment insurance and decreasing its duration, which would not come at any cost to the intangible values that you have been defending.
 

Posted: Thurs, Mar 13, 2003 at 13:45:18 (EST)
Original: Tues, Mar 11, 2003 at 19:55:14 (EST)
Posted by: Erik Rauch  Recipient: Bobby
Email Address: erauch @ ai.mit.edu

> It's not true among academic economists.
> They are concerned with that which is hard to model, and
> they find a way to model it or to discuss it intelligently. Do
> you know who Amartya Sen is?

Yes, but again, I am talking specifically about rich countries, where the real benefits of increased income are getting smaller and smaller, and so it makes sense to forego some increase in income in favor of non-monetary benefits.

> It is consumption that beings brings pleasure. My point was
> that yearly forgone consumption is the cost of these policies.
> And that cost grows exponentially and that is a huge cost to
> any policy that creates DWL.

The difference doesn't grow exponentially no matter how you measure it. The difference in any given year is a constant percentage of GDP. You seem to be counting the cumulative difference between the consumption that would have occurred with and without the policy, but this doesn't grow exponentially either. What about the foregone value of a more equitable system? That will add up too, even if it is harder to assign a dollar equivalent.

This illustrates my point. You consider the foregone consumption to be a 'huge' cost, while ignoring things that don't show up as consumption or GDP.

> There are theoretically and empirically sound ways of
> decreasing the natural rate of unemployment such as
> reforming unemployment insurance and decreasing its
> duration, which would not come at any cost to the intangible
> values that you have been defending.

I don't regard simply cutting people off unemployment benefits after a few months as having no downside.
 

Posted: Thurs, Mar 13, 2003 at 17:06:30 (EST)
Original: Thurs, Mar 13, 2003 at 13:45:18 (EST)
Posted by: Bobby  Recipient: Erik Rauch
Email Address: robert @ pkarchive.org

People do NOT consume percentages of GDP. They consume real dollars, PPP, or whatever measure you use. In real PPP, dollars or whatever, any policy that brings you to a new balanced growth path, the difference between the per capita GDP, social surplus, or whatever you could have achieved without the policy grows exponentially. It DOES grow exponentially, and you have not disproved this in any way. If the difference between two economies one at level G and the other at level S on balanced growth paths is initially
(G - S)
and if each economy grows at g the differential is always
(G(1 g)^t - S(1 g)^t)
or
(G - S)(1 g)^t
which shows that the differential grows at g or exponentially
What is important to people's pleasure is (G-S) or real PPP or real dollars or the vertical distance as you said before. That is what is important, not the time between levels of GDP, and not percentages of GDP. Again you have forgotten that forgone GDP also implies forgone social surplus, so I am not talking only about figures that show up in GDP figures. Remember that consumption, social surplus, etc. are flows that are created each year. I am looking at the forgone flow each year. And the forgone amount of those flows increases exponentially over time.

I am not sure how equity problems grow exponentially, but these equity can be taken care of through proper taxes and subsidies that redistribute from winners to losers instead of these command and control regulations that inhibit not only create static DWL but inhibit growth. The income tax and modified EITC are what I have in mind. Though these taxes and subsidies do have static DWL effects you have design them so that they do not inhibit growth and therefore won't have exponentially growing forgone social surplus.

I never said unemployment insurance reform has no downside, but you must do SOMETHING that decreases the reservation wage of those unemployed on unemployment insurance relative to labor market wages, and this workweek policy does not address that problem. I am not sure how indefinite and poorly designed unemployment insurance increases the equity or justice any of these intangible values. It does create unemployment. I'm fine with increasing the duration temporarily during a recession but that's it.

It is not clear to me how your comment about benefits of income getting smaller and smaller addresses my rebuttal of your untrue accusations upon economists. Nevertheless there is much good that can be done with increased income besides fattening well-healed students or upper-middle-classers, like you and me, for which diminishing returns is a concern and besides it being wasted on dubious programs like this workweek plan. It can be redistributed to the poor in France by subsidizing their work hours or to the poor in other parts of the world, or to correct a market failure through some subsidy. Increasing aggregate income is VERY important for the government to be able to fulfill its future obligations as well. Despite our own personal material comfort at Yale and MIT (though I'm sure MIT is much nicer) resources are still scarce in this world.
 

Posted: Thurs, Mar 13, 2003 at 17:56:51 (EST)
Original: Thurs, Mar 13, 2003 at 17:06:30 (EST)
Posted by: Bobby  Recipient: Bobby
Email Address: robert @ pkarchive

Sorry this should read If the difference between two economies one at level G and the other at level S on balanced growth paths is initially
(G - S)
and if each economy grows at g the differential is always
(G(1 g)^t - S(1 g)^t)
or
(G - S)(1 g)^t
which shows that the differential grows at g or exponentially

Sorry I see what the problem is. What I said was 'And that cost grows exponentially and that is a huge cost to any policy that creates DWL' This is a typo. It should say any policy that puts the economy on a worse balanced growth path.
 

Posted: Thurs, Mar 20, 2003 at 19:37:46 (EST)

Original: Thurs, Mar 13, 2003 at 17:06:30 (EST)
Posted by: Erik Rauch  Recipient: Bobby
Email Address: erauch @ ai.mit.edu

> The income tax and modified EITC are what I have
> in mind. Though these taxes and subsidies do have
> static DWL effects you have design them so that they
> do not inhibit growth and therefore won't have
> exponentially growing forgone social surplus.

Wait a minute - you are employing the same argument that I did. If they have any static effect, then they will have an 'exponentially growing forgone social surplus' as you put it, although I think it is wrong to worry about a small constant fraction of an exponentially increasing pie. Of course this doesn't matter much, because you have merely traded a few years' worth of growth for the immense benefits of more equality.
 

Posted: Wed, Mar 19, 2003 at 19:53:58 (EST)
Original: Thurs, Mar 13, 2003 at 17:06:30 (EST)
Posted by: Erik Rauch  Recipient: Bobby
Email Address: erauch @ ai.mit.edu

> In real PPP, dollars or whatever, any policy that brings you