22 February 2012

Does Increased Productivity Decrease Manufacturing Jobs?

I was intrigued to see an argument being made in a recent Brookings Institute report (Helper et al. here in PDF) that productivity gains do not lead to job losses in manufacturing. This post explains why I think that this argument is not well characterized and probably just wrong.

Helper et al. make the following argument (pp. 9-10):
Some argue that strong productivity growth has caused much of America’s manufacturing job loss, especially in the last decade. This theory, which contends that technology is replacing workers, stems from the observation that apparent productivity gains have coincided with manufacturing job loss in the 1990’s and 2000’s. Yet there is no economic reason why increased productivity must lead to job loss.
The first thing to note is that Brookings is talking about overall productivity gains, and not labor productivity. The different is important. Labor productivity according to the BLS "is the ratio of the output of goods and services to the labor hours devoted to the production of that output." Overall ("multi-factor") productivity "relates output to a combination of inputs used in the production of that output, such as labor and capital or capital, labor, energy, materials, and purchased business services (KLEMS). Capital includes equipment, structures, inventories, and land." The distinction is crucial to understanding employment changes in manufacturing.
There are a lot of moving parts in multifactor productivity beyond labor. So in any analysisa good first place to start, fairly obviously, is the relationship of labor productivity and jobs. The above graph shows the relationship of changes in labor productivity with changes in employment for manufacturing, according to data from the BLS (with 2005 = 100 for each time series).

It shows an extremely close relationship between labor productivity and employment, one that is even stronger on a one-year lagged basis (as companies fully exploit labor productivity gains). This relationship should not be at all surprising since labor productivity is defined in terms of hours worked, which has a direct relationship with employment.

So the first order conclusion necessarily must be that gains in labor productivity in manufacturing necessarily lead to losses in jobs. And it is here that Brookings gets a bit confused with its focus on total productivity:
Even though a productivity increase means that fewer workers are needed to produce a given quantity of output, the productivity increase also allows product prices to be lower, increasing the size of the product market. The bigger market means that firms will need to hire more workers. The additional hiring needed to produce for a bigger product market usually offsets the initial labor-saving impact of the productivity increase. Therefore, the overall impact of a productivity increase is usually to expand employment rather than reduce it.
To see where Brookings has erred, think about the mathematics here -- If a market for a good expands that means that more output is required. Additional labor needed to meet that demand is an input. If the magnitude of the change in the input is equivalent to the magnitude of the change in output, then there has in fact been no increase in labor productivity, which is defined as an ability to produce more with fewer hours on the job. If labor productivity is improving then the inexorable result will in the long run be a loss of jobs, even in an expanding market.  In the short term, market expansion can offset job losses, such as we are currently experiencing in the US, but only temporarily. In short, with respect to jobs, simple math tells us that market expansion cannot defeat labor productivity. Just look at agriculture for an example of this dynamic (e.g., programs such as the Food for Peace effort of the 1950s were designed to expand markets in the name of aid and diplomacy).

Now, with respect to total productivity it is certainly possible that changes in other inputs can work to offset gains in labor productivity, but only at the expense of overall productivity (i.e., if labor productivity improves then KEMS productivity must degrade).  Similarly, total productivity gains can occur even if labor productivity degrades or is held constant (i.e., via improvements in KEMS productivity). Thus, there is nothing inconsistent with overall gains in productivity and constant or even increasing employment, but such an outcome does require that labor productivity gains are smaller than the combined effects of market expansion and degradation in other sources of productivity gains. This is a tough ask: The case of Germany suggests that even aggressive policies to influence labor productivity (which in some cases degrade labor productivity) .

Does increased productivity in the manufacturing sector result in a loss of jobs? If those productivity gains are in labor then the answer must be yes.


  1. .

    Increased manufacturing productivity leads to fewer manufacturing jobs. The very same was true in agriculture.

    However, that simply means that there will be more jobs in marketing, healthcare, hospitality, etc. For example, the only way high cost services like physical therapy can be widely afforded is if the cost of food and manufactures is low.


  2. If you go a long way back to the day of Rockefeller and Standard Oil, they way Rockefeller took over the industry was through productivity gains. He could make product much cheaper than his competition, much cheaper. In other words, the productivity of his company was much better and it was a darn good thing considering the growth of the oil industry. But low cost fuel also fed into the success of his industrial counterparts, Henry Ford, to make low cost cars that dramatically expanded the transportation market. Low cost transportation was essential in the growth of the suburbs which spawned drive in restaraunts, recreational vehicles, low cost travel, the list goes on and on. If you drive down the cost of something that people want to do based on productivity improvements, other people will develop new products and to help those who were successful spend the extra money in their pockets and that will create new opportunity (and even more jobs). Look at the other side of the coin, find an industry whose product people feel they must have but productivity has declined over the years. (Two come to mind off the top of my head.) Ask a hundred consumers of those products if you think we are better off because the prices are higher.

  3. Good analysis, Roger.

    When you have increased productivity, whether labor productivity or otherwise, the costs of the good or service decline. This puts more money in the pockets of consumers, and they can and do spend more $, creating new jobs, as the first commenter suggests.

    This increased expenditure may or may not lead to a net increase in jobs in the US. It is a matter of empirical study.

    If more of that increased spending occurs for goods or services produced abroad, the first order effect may be less jobs in the US, but there is also another round of cost reduction (because the goods produced abroad wouldn't be purchased unless cheaper than the made-in-USA brands).

    How it all turns out is again a matter for empirical study. For example, I read a which found that a Walmart in a poor part of a big city would save a poor family $2000 on purchases that would have otherwise cost $8000 per year before the Walmart arrived. $2000 in the pockets of the poorest among us seems a pretty good social policy. But I don't know if there would be net job creation.

  4. In the LONG run, and globally, if the efficiency of production (agricultural, manufacturing, etc.) continues, the workers displaced by such new technology will have lost THEIR purchasing power and will not be able to be consumers of even that more efficient production.

    Only IF – in a VERY short period – at least equal numbers of new (service?) jobs, that promise to increase the quality of life (QOL) of THEIR potential consumers, will there be enough potential consumers of the products of the highly efficient 'manufacturers', to avoid disaster!

    This is NOT an empirical matter! It is a matter of logic.

    There is no reason to believe that such matching desirable change will automatically occur as a result of the motivations for GROWTH IN PRODUCTIVITY.

    Only if something like a replacement of that motivation, with a motivation for continuous GROWTH in QOL, might there be a solution to this problem.

    But this is extremely unlikely under the current dogmatic paradigms of either economic supply-side or demand-side theories!

  5. I worked my whole carer for a manufacturing company. When we invested in a equipment to save costs we expected to pocket those savings. There were the ROI that justified it. Companies such as this resist with all they've got to lower prices. The only thing that lowers prices of consumer goods is competition. We sold to Wal-Mart. It was our largest customer. They pit one supplier against another and threaten you with the loss of shelf space. Ask any sales person who has visited Bentonville, AR about the deliberately spartan conditions when meeting a purchasing agent.

  6. You are a much better qualified economist than I, who has only a first degree in (Mathematical) Economics (plus post-grad degrees in Science and Business). I must, however, take issue with your mathematics. Ceteris paribus, whether an increase/decrease in labour productivity results in an decrease/increase in anything depends (as does everything in economics) on the elasticities. Surely?

  7. -6-Ir'Rational

    Well, economics is not among my degrees, so you've got me there ;-)

    If labor productivity is defined as $s-of-output/hours-worked then it increases only by increasing the numerator and/or decreasing the denominator. Brookings is arguing that an increase in the numerator will lead to an increase in the denominator. But if labor productivity improves, then this means that the jobs/output will necessarily decrease.

    So while manufacturing can still play a big role in the aggregate economy, jobs as a % of total jobs will go down, which is what has been the case.

    I'd welcome your thoughts, Thanks!

  8. re 6

    Wouldn't your issue with elasticities be answered with the knowledge that firms will have taken elasticity into account. If the increased productivity did not have a positive return on investment then it would not have been chosen. So the case to be considered in the one in which there is a positive return.

    I have no qualifications whatsoever in economics and am merely asking the question

  9. Labor productivity is a measure of the amount of money that is going to workers. If labor productivity increases then the amount of money going to workers will decrease. The cost of goods goes down by the amount of decrease in labor costs. The income of workers goes down by the amount of the decrease in labor costs. The total amount to buy goods remains the same as does the total cost of money available to buy goods. The firms that lower prices have advantages against firms in their own market which cannot take advantage of the new means of cost savings and as well firms outside of their markets who are similarly unable to benefit.

    So it seems like a zero sum game in labor productivity for the whole economy. Is there any credibility to this analysis. It seems to be to the same level as that of Brookings analysis

  10. -9-djvjbsl

    The analysis with respect to this post is on one sector of the economy. Consider that in 1900 agriculture was something like 40% of jobs and today is <1%. Improved labor productivity was one factor in the diversification of the economy.

  11. re 10 - Roger Pielke Jr

    I think that what I am saying is compatible with what I think that you are saying. Increasing labor productivity can shift jobs from one sector of the economy to another but it cannot create new jobs

    More accurately I think it shifts money from one sector to the other and the same number of jobs would be created would remain the same if there is an assumption that the wages for a job remains the same. However, if wage costs per employee are reduced (e.g. jobs shifted to China), there may be more people employed but each for less money. No money is created or destroyed. It is just rearranged.

    I think that this works to the same level as the Brookings argument so it is probably just showing that that level of argument is inadequate.

  12. Roger was it you who linked recently to the excellent Atlantic article 'Making it in America'?

    There was a cracking quote early on about manufacturing productivity:

    "There’s a joke in cotton country that a modern textile mill employs only a man and a dog. The man is there to feed the dog, and the dog is there to keep the man away from the machines."

  13. It depends.

    If the productivity of widget makers doubles, cutting cost enough to increase demand for widgets tenfold, then the number of widget makers will increase fivefold. If productivity doubles again, increasing demand fourfold, employment will double again.
    But as the market for widgets matures, it'll reach the point where doubling productivity will only double demand, increasing employment only onefold. After that, increasing productivity won't increase demand enough to sustain employment, much less increase it.
    And after demand is saturated, employment can only decrease.

  14. -13-Bill

    Thanks, this is a good example why I argue: "In the short term, market expansion can offset job losses, such as we are currently experiencing in the US, but only temporarily. "

  15. When considering the effects of productivity improvements on labor, I’m bemused by the incentives facing companies in the same highly competitive industry. Take GM and Toyota, where demand for the latter’s product is higher than for the former in some segments. The latter has had every incentive to boost productivity here in the US to grow its market share and can do so in part because it is not burdened by union work rules and other labor-productivity-killing obstacles. That most Toyota factories can sell their growing output allows it to use all the tricks (more subcontractor assembly, readily shifting workers among different workstations, etc.) and still maintain employment levels.

    GM, on the other hand, has had fewer incentives to boost labor productivity because it must continue to pay union workers whether they’re on the line or in a study hall. That’s the main reason in the 1990s that GM has never seriously tried to get down to Toyota’s 17 man-hours per vehicle: it would receive little benefit if it did so. (To be sure, as the largest US manufacturer, it’s harder for GM to expand its market share, quality, price, and desirability aside.)

    Reduced productivity such as the sky-rocketing price of natural gas in the mid-2000s resulted in dramatic employment reductions in the US as DuPont and other chemical producers that use natgas as a manufacturing input shifted production from the US to Asia, primarily Indonesia.

    But look at the turnaround now that fracking has caused the price of natgas to tumble: employment in related industries is increasing, another case of increased multi-factor productivity. Per today’s Wall Street Journal, French steel company Vallourec & Mannesmann Holdings, Inc. began construction on a new $650 million plant to make steel tubes for the hydraulic fracking industry. Where are the 400 workers building the plant? In the heart of the rust belt, Youngstown, Ohio, and you can be sure there will be plenty of workers available as soon as the presumably state-of-the-art plant opens for business.

    Whether increasing multi-factor productivity reduces overall employment is an important question, but the answer is elusive.

  16. Another couple of points:
    -If productivity increasing technology is broadly applicable across industries, then the economy's ability to absorb displaced workers will be heavily muted in the short and medium term (and their reduced purchasing power partially offsets the productivity gains).

    -It is much easier to think of uses for technology in existing industries than it is to think of entirely new uses.

  17. So I suppose the Brookings theory is that some workers are displaced by the increase and productivity and lose their income. However, other workers will see lower prices and so be better off. This is essentially a transfer of wealth from the displaced workers to the others. However the increased purchasing power of the non-displaced workers will allow them to buy more goods thus stirring growth in their own industries and as well in new start up industries that spring up to take advantage of this new wealth.

    The displaced workers will either be supported by new money in the form of income support programs from the government or borrow against their future earnings. Thus new money will be created which increases the total beyond that that existed before. That is there will be the existing wealth, now redistributed, plus the income support and borrowings for future earnings by the displaced workers. This new money can support the development of new markets which can take advantage of the labor of the displaced workers.

    I suppose that this is a description of the Brookings argument. The wealth released by the increase in productivity will be used to fund growth and the emergence of new markets. I can see how wealth is created and the standard of living raised but there appears to be no guarantee that any new jobs will be created. Technological innovation to create the possibility of new markets and to support increased labor productivity could keep this process growing indefinitely

  18. -17-djvjbsl

    Thanks, though I don't think that is quite the Brookings argument, as they are focusing within the manufacturing sector. Empirically and historically I think that it is indisputable that increasing economy-wide productivity leads to new jobs. The question is whether manufacturing is somehow a special industry compared to the rest. I remain unconvinced. Thanks!

  19. Roger,

    I'm not an economist so I'll take everything you said at face value, it does seem to jive with everything I learned as an undergrad.

    I have a question though, tangent to this topic and maybe take it on as a topic for a future post, but you post brought it to mind, which is about Social Credit theory, something I have only recently even heard of [certainly never mentioned in undergrad macro].

    Social Credit theory SEEMS appealing because it is the only economic theory that really tries to take account of and embrace exactly this inevitability of job loss due to productivity gains do to improvements in manufacturing & etc.

    Appealing, but does it actually work? Can it work? Experts seem to differ wildly in their opinions.


  20. The very best example of increased productivity increasing the number of jobs, along with revenue and profits, is McDonald’s 1976 introduction of its breakfast menu. Adding five hours of operation to its restaurant chains increased raw materials, labor, and utilities costs, but provided increased revenue that could be applied against fixed plant and equipment costs. Overall Mickey D’s and its franchisees increased profits significantly while increasing employment.

    The only downside is cultural: does the McDonald’s at a given location switch from breakfast to lunch service at 10:30 AM or 11:00 AM? Now there’s a candidate for Brookings’s researchers if there ever was one, no?

    Which prompts me to assert that any study of American ingenuity / innovation is incomplete without at least an honorable mention of the Egg McMuffin.

  21. re 18

    Just to extend the supposed Brookings argument described in 17. According to 17, the result of productivity improvements is generate new wealth that will be used to open new markets by encouraging entrepreneurialism and new investment. That is, people will borrow from the future to create the new wealth. The success of this will encourage optimism that things get better as technology progresses and this will promote the optimism that will encourage entrepreneurialism in itself. A culture of risk taking will be slected for in which people will take the risk to open new markets and thus create employment.

    So if this is so, the Brookings argument is that the wealth set free by productivity improvements encourages the optimism that induces the risk takings that create new jobs. A positive feedback is set up which probably creates the expansion portion of the business cycle.

  22. Roger,

    Somehow I'm not following your logic. You say "If labor productivity is improving then the inexorable result will in the long run be a loss of jobs, even in an expanding market."

    Net jobs should increase if the market growth is greater than labor productivity growth, right? No necessary connection between labor productivity and jobs can be inferred without knowing the nature of the market expansion. The Brookings argument, then, where they assert (market growth) > (labor productivity growth) would seem to be an empirical question and not one of logic alone (i.e. there is no inherent logical flaw in the their argument, as you seem to assert). Or have I missed something?

    In any case, doesn't the empirical record support Brookings? Over long periods of time, hasn't U.S. labor productivity AND employment (in terms of total jobs) been increasing, as you almost point out ("it is indisputable that increasing economy-wide productivity leads to new jobs.")?

  23. -22-Brian

    Thanks ... the Brookings argument is not about the economy as a whole, but about the manufacturing sector specifically. What holds for the entire economy does not necessarily hold for a particular sector.

    Empirically, the data does not support Brookings. Look at the data here on manufacturing jobs:

    early growth, long-time stagnation, rapid decline

  24. Roger,

    Yes, but your counter-argument doesn't reference manufacturing (or any sector) explicitly, so it would have to hold for many sectors if it is correct. Or have you assumed something that I missed that makes it specific to manufacturing? This is where I lose you.

    Regarding the graph you've linked, I think it's hard to attribute the declines (only since 2001) to labor productivity increases. Isn't much of that decline related to offshoring? The two dramatic falloffs hardly seem consistent with productivity gains, which I would expect to be more steady and slow.

  25. Pro Pielke

    Is your point in this that manufacturing jobs are jobs like any other? That is that the concern over the loss of manufacturing does not take into account that this is a natural result of the technology of manufacturing and that trying to stop it will be futile? Or , to the same point, The loss of jobs in the specific category of manufacturing is the cause of the increase in jobs in other sectors

  26. -24-Brian

    Thanks, have another look at the scatter plots in this post, they show the relationship of labor productivity and employment, hard to argue that they are not related (in fact, the relationship is definitional!)



  27. Does increased productivity result in a loss of jobs? If those productivity gains are in labor then the answer must be yes.

    This seems to ignore the history of the world for the last several centuries at least where productivity increases and increased employment have gone hand in hand.

    Who you gonna believe, that graph or your own lying eyes?

  28. -27-AAA

    Thanks, I must have been too opaque, so I have edited the last sentence ... better? ;-)


  29. Hello,

    I’m glad to see all the interest in our report! I have a few clarifications that may resolve much of the discussion here:

    First, our report does in fact use labor productivity data, not multi-factor productivity, as we state pretty clearly in the report (pages 8 and 10). We agree with the author of this post: multi-factor productivity would be the wrong measure to use. BLS has a description of their labor productivity data here: http://www.bls.gov/lpc/.

    As for the two figures in the post, I agree with Brian in post 24 above when he says the figures don’t tell us much about causality between productivity change and employment change. It’s important not to mistake correlation for causality in the first place, especially when the comparison only considers one decade. But the more important issue (explained in more detail in Box 1 on page 7 of our report) is that current BLS data is actually not capable of distinguishing between cost reductions from outsourcing and actual, domestic labor productivity gains. So I think Brian is exactly right when he says a lot of what’s being captured in BLS data is actually offshoring, not productivity gains. The finding that offshoring is strongly correlated to US manufacturing employment loss is logical indeed!

    Even if you trust the BLS data, which overstates mfg labor productivity growth, it’s a tough sell to chalk the massive US employment loss up to productivity gains. Labor productivity growth appears to have been fairly consistent between the 90’s and the past decade, while mfg employment loss accelerated ten times over from one decade to the next. Productivity growth just doesn’t fit as the independent variable. Rob Scott makes a similar argument here (http://www.epi.org/publication/webfeatures_snapshots_20070221/), noting also that employment loss does move with a large drop in US mfg output. I think this is further indication that BLS data is confusing offshoring (i.e. decline in output) with labor productivity.

    Finally, I would say that everyone here is correct to raise the question: if productivity growth doesn’t inevitably lead to employment loss, why doesn’t it? The answer is that economic reality doesn’t make sense if you think of one nation’s economy as a zero-sum game. But consider that:
    a) one single actor (a country, a firm, a region) can gain jobs in an industry by winning them away from peer actors even if the global employment in that industry is stagnant or declining. Productivity is a major method by which one nation, region, or firm can win this competition and increase their employment by becoming more productive and thereby decreasing costs (at the expense of employment throughout the rest of the industry).

    And, b) so far, humans have never stopped consuming more goods in the long term, meaning even global manufacturing is not a zero-sum game. This isn’t the case for all goods – if each citizen of a poor country eats 7 bananas per week, and then over time the country’s wealth quadruples, they won’t start eating 28 bananas per week. But they could reasonably be expected to spent 4x more on computers, cars, and other higher-end products. So as productivity increases and help countries grow their standard of living, markets grow, as does employment (as long as you’re manufacturing something people in your country or in other countries want more of). It’s a fairly non-controversial statement to say that many industrialized countries have expanded their manufacturing sectors in this way.

    So if people somewhere are consuming more of a product, productivity growth just means you’re growing your numerator (output) more quickly than you’re growing your denominator (hours worked). I think the second blurb from our paper that Roger cites above explains this concept well.

    Anyway I’d encourage everyone to take a look at the paper if you haven’t, and we really appreciate the interest in this important topic.

    Tim Krueger

  30. -29-Tim Kreuger

    Thanks for stopping by ... just a few quick comments in response: some of what you write sounds like a bit of dissatisfaction with the definition of "labor productivity."

    In fact you've emphasized "domestic" labor productivity versus outsourcing ... but doesn't this come out in the wash if we are discussing multi-factor productivity?

    If more MF productivity is a good thing, then we should welcome a chance to outsource parts of the supply chain, even if the result is a loss of domestic jobs, right?

    Also, can you point to industrialized countries that have expanded jobs in the manufacturing sector?

    Thanks again, and kudos on issuing a report worth both reading and responding to ;-)

  31. Roger,

    1) It's a bit hard to understand the two graphs at the top of the post since the axes have no labels. I don't know what the numbers mean.

    2) The relationship you are arguing for (increased labor productivity --> declining manufacturing jobs) can't possibly be definitional. After all, manufacturing jobs increased through the 1930s - 1960s while labor productivity presumably increased. The decline in the last decade must be due to other considerations and not labor productivity per se.

    3) Outsourcing is indeed a very good thing. It may be painful in the short term, but it leads to a higher standard of living for all in the long run.

    4) What Mr. Krueger said. ;)

  32. -31-Brian

    Fair points ;-)

    1. 2005 = 100 for each time series
    2. Labor productivity = output ($)/hours worked

    It is indeed definitional

    Your 3. and 4. are incompatible


  33. ========
    3) Outsourcing is indeed a very good thing. It may be painful in the short term, but it leads to a higher standard of living for all in the long run.

    Another way to look at the issue of outsourcing is to consider the factor of "deskilling". Labor productivity is being increased now by taking the skills required for manufacturing and placing them in the computers and robots that now work the line. Cost savings are achieved by "right sizing" and by requiring only lower skilled lower paid workers who can be hired and fired to achieve the lowest cost.

    The increased wealth being generated by the technologies is not being shared with the manufacturing class but is being used for executive compensation and financial transactions. So income inequality increases and wealth that was used to raise the standard of living of people with modest incomes is now lying dead in Park Avenue real estate etc.

    So new wealth is being generated but whether that increase the quality of life for all is highly questionable

  34. Roger,

    I agree that labor productivity = output/hours worked is definitional. That's trivial. What is not definitional is your claim (if I understand you correctly) that increased labor productivity in a given sector necessarily leads to fewer jobs in that sector. Output must be considered too.

    I don't see how statements 3 and 4 are incompatible. Please explain why you think so.


    In a competitive environment (assumed if "right sizing" is happening), cost savings are passed on to the consumer. This causes the standard of living to rise for all--they have more money to buy other things.

    The free markey is all about efficient allocation of resources. It is wasteful for all if more is spent on making a product than is necessary. It is harmful to human dignity to have people doing jobs that can be done by machines. Technology allows people to work in new areas better suited to their unique abilities. It's a benefit to all, though it can cause painful displacement in the short term.

    I don't accept the notion of a "manufacturing class." People may work in manufacturing, but that does not define who they are.

    Finally, the idea that the wealthy are squirreling away their money and depriving the middle class of upward mobility is a complete fantasy. Unless people are putting their money in a shoebox under the bed, it is being passed around in commercial exchanges. As long as wealth is circulating, it gives everyone a chance at getting a piece of it.

  35. -34-Brian

    Thanks ... from 2001 to 2010 (BEA data) manufacturing output increased by 27% (current $) or 3% (after adjusting for inflation). Over the same time period labor productivity improved by about 36%.

    For output to have (over)compensated for the effects of would have required growth in output of >36% from 2001 output. Overall GDP grew over that period by about 25%. So the manufacturing sector would have had to have grown faster than GDP, making it a larger part of the economy. Such mathematics are not sustainable as under such a circumstance eventually manufacturing will reach 100% of the economy, and labor productivity will once again lead to decreasing jobs.

    Now is it possible that manufacturing output could grow fast enough to offset gains in labor productivity but not as fast as GDP? (That is GDP change > manufacturing output change > labor productivity change). Theoretically, yes. But with labor productivity improving by 36% over 2001 to 2010 and GDP growth in manufacturing at 3% (constant $) there is no such sweet spot -- not even close.

    Policy makers can try to hit this sweet spot by modulating labor productivity (by reducing it, a la Germany) and artificially trying to boost output (a la cash for clunkers) but the target has proven a hard one to hit except for in the very short term and in special cases (I tried to get at this in the second to last paragraph above, worth expanding on in another post).

    Thanks again for the comments and a chance to work through some of these issues in a bit more detail.

  36. re 34

    Brian writes

    Finally, the idea that the wealthy are squirreling away their money and depriving the middle class of upward mobility is a complete fantasy. Unless people are putting their money in a shoebox under the bed, it is being passed around in commercial exchanges. As long as wealth is circulating, it gives everyone a chance at getting a piece of it.

    Yes, money circulates but in its circulation the proportions by which it is divided remain the same. A decreased proportion goes to the workers and and an increased proportion is used for executive compensation and and for financial transactions. Increased productivity is about reducing the cost per worker. The money that it generates must come from somewhere and currently that comes from deskilling the work force and paying lower wages.

    Money circulates and is divided. There is no certainty that this division will create an affluent middle class of industrial workers. That was the contingent outcome of circumstance and government policy in the mid-20th century. Conditions are changing and the available jobs do not have the value added to justify the wage levels that existed previously for the numbers that were previously supported. There are fewer higher skilled jobs interacting with the robots and more lower skilled jobs for which robots are not yet cheap enough to displace the human worker.

  37. This article from the Atlantic is of interest to this topic. it uses the history of the American whaling industry to illustrate what is happening in manufacturing.

    The Spectacular Rise and Fall of U.S. Whaling: An Innovation Story


  38. @ Roger 28

    It was clearer before. In the jobs world the only thing that really matters is the total number of net wealth producing jobs. How they get classified is a matter of classification gamesmanship and is in some ways as useful as debating the number of angels that can dance on the head of a pin.

    Example 1. Manufacturer A quits using 50 workers at advertising agency B and instead hires 50 workers to do advertising in house. Note the loss in manufacturing labor productivity as 50 more workers are employed for the same output.

    Example 2. Manufacturer C quits using 50 in house workers for advertising and instead uses 50 workers at advertising agency B (who are happy for the work having just lost the A contract). Note the gain in manufacturing labor productivity as 50 fewer workers are employed for the same output.

    When the fastest way for a manufacturer to boost manufacturing labor productivity is to contract out for services the any discussion of "manufacturing labor productivity" is a waste of time.


  39. I'm not sure you have answered my objections. If the demand (unusual, I admit - and even more unusual, the supply) curves have negative slope (the usual example given is Rolex watches [although the British benefits system comes close])it might lead to a different result. In other words, that casts doubt upon your thesis. I don't know (you would have to prove the relationships you assume) whether your thesis holds.
    It's attractive - but not justifiable! (I think - the extreme case demonstrates &c.)

  40. -35- Roger,

    Policy makers can try to hit this sweet spot...

    "Policy makers" would do well to get out of the way and stop trying to hit any imaginary sweet spot.

  41. -36- dljvjbsl
    You wrote "The money that it generates must come from somewhere..."
    This assumes that total wealth is static; that one person's enrichment is at the cost of others. I don't think this is correct. Increased productivity frees resources to increase total wealth.

  42. -38-AAA

    Thanks, and I agree ... however much of the current debate is about a sector characterized as "manufacturing" under the NAICS.

  43. Re 41
    Howard wrote:

    -36- dljvjbsl
    You wrote "The money that it generates must come from somewhere..."
    This assumes that total wealth is static; that one person's enrichment is at the cost of others. I don't think this is correct. Increased productivity frees resources to increase total wealth.




    For a generation, middle class and lower incomes have been stagnant in real inflation adjusted terms while those of upper incomes rose dramatically. Productivity improvements are now resulting in rising inequality. The total wealth is increasing but that wealth is being reserved for executive compensation and bankers' income.

    I haven't seen any research on this but to me a major factor must be that the new technologies are creating their economies by deskilling,

    I read an account of a newspaper columnist who writes for the Daily Telegraph but lives in New York. She goes to parties and commmonly hears remarks at parties such as "Remember 20 is just 10". That is an income of 20 million a year is really just 10 million after taxes. She wrote of the insularity of these people who live in a different reality. Is some investment banker really worth 20 million a year? The research shows that star CEOs do not fare very well if they move to a different company. Were they paid 10s of millions year at their previous company because of their unique skills or blind luck.

  44. My thoughts in response to comment 30 above:

    In re productivity gains: it sounds like we're just talking about whether it's better to lower prices via offshoring or via higher value-added per domestic productivity. I think the answer depends on your scope of concern, since the benefits of lowering prices will be captured by different workers and employers depending on how you do it. It's better for low-cost economies and their workers if we lower prices by offshoring; it's better for US innovation and long-term competitiveness in high-end markets/future technology cycles if we increase productivity through higher value-added at home instead of simply cutting costs and offshoring (we present our evidence for this argument in section A2 of our report). Overall, we're assuming that US public policy should primarily be concerned with achieving goals that are in America's long-term interest, but one were to disagree with this starting premise I guess they could make the argument that offshoring is the better way to increase multi-factor productivity and lower prices.

    Also we think it's best if at least some of the benefits of productivity gains accrue to workers and are not just captured by firm owners. We think a key goal of US economic policy should be sustaining a middle class. This is most likely to happen if the method of achieving productivity gains is growing value-added instead of cutting costs (as dljvjbsl explains in post 33 above).

    In re advanced nations growing mfg employment:
    According to BLS Intl Labor Comparisons data (http://www.bls.gov/fls/flscomparelf/employment.htm#table2_4):
    -Canada had consistent mfg employment growth from 1993-2004, and net growth from 1993 as far as 2009
    -Germany retained constant mfg employment from 1997-2008, which is a major accomplishment considering national trends during that period
    -Italy had mfg employment growth from 1987-2003, with net growth from 1987 to 2008
    -Japan grew their manufacturing employment from 1978-1992
    -Netherlands has growth from 1983-2002
    -Spain also grew their hours worked form 1990-2000 according to this BLS data: http://www.bls.gov/news.release/prod4.t01.htm

    So, certainly, mfg decline has happened in many places during certain time periods, but mfg growth is also possible and happens in certain places during certain time periods.

    And some other countries, like Denmark, have instituted effective policies to slow the decline of their manufacturing sectors. This is even more significant given that countries like Denmark (and I would assume most northern European countries) have had slower population growth than the US, making their comparatively less harsh mfg declines all the more impressive. So while the general trend has been for manufacturing employment to decline in advanced industrial countries, nowhere has it been as devastating as the US and the UK (in sheer numbers and % of employment, I believe). Thus, if you agree with our arguments for why there's value in keeping a manufacturing sector, it still seems clear to me that the US has some important lessons to learn from other advanced economies.

    I think Canada's experience - Canada whose economy, manufacturing and otherwise, is highly intertwined with America's - really shows that they must have been making some decisions differently than in the US. Canada lost 287k mfg jobs, or 14% of its mfg employment from 1979-2010, while the US lost 8.38 million (37%) during the same time period.

  45. -44-Tim Kreuger

    Thanks ... Canada is indeed a very good point of comparison.

    Here is data on Canadian manufacturing jobs:

    Which today is lower than at any point since 1963 Indeed the data you link to shows 2010 as the lowest year in the data series).

    Also, Canada has had an enormous problem with low productivity growth in manufacturing:

    Official data show a degradation in productivity:

    The same data that you view as a virtue, some analysts view as a vice:
    “The conundrum is that slow productivity growth has not only been a fact of Canadian life for the past three decades and continues to worsen, but it has also been impervious to numerous policy initiatives intended to improve it”

    Unlike the US, Canadian manufacturing output has been declining:

    Canada's inability to secure consistent increases in productivity mean that not only is it losing jobs, but it is also losing output -- not a good model to emulate.


  46. "Does increased productivity in the manufacturing sector result in a loss of jobs? If those productivity gains are in labor then the answer must be yes."

    I think a coupled question is the "Does increased productivity in the manufacturing sector result in growing income disparity?" As each of those higher productivity jobs has both higher skill requirements (and thus potentially draws from a narrower population), and has a higher marginal contribution, the result is going to be a higher relative wage compared to sectors without those productivity increases.

    Given that the those unskilled, untrained, or incapable of filling those jobs increases the pool of unskilled *and* unemployed, they unskilled market becomes saturated, further driving down wages.

    That's all theory and supposition of course... but I pose it as a related question.

  47. So Canada may or may not have some lessons for the US to learn from, but the discussion seems to have taken a turn now:

    "Canada's inability to secure consistent increases in productivity mean that not only is it losing jobs, but it is also losing output..."

    It sounds like maybe we're agreed that productivity gains are important for creating manufacturing jobs, not just replacing them?

    Anyway I think the main point is that we don't actually know what US labor productivity is, due to the major flaws in the BLS data. For all we know it could be worse than Canada's (unlikely but a possibility). Susan Houseman's work shows that actual US labor productivity gains in mfg have probably been quite low, and even lower if you look outside computers/electronics. Nordhaus' work suggests US mfg jobs loss would have been worse if productivity gains had been lower in the last 2 decades.

  48. The issue that Canada has with lack of productivity improvement is commonly attributed here to the policy of the low dollar. The Canadian dollar was allowed to depreciate against the US dollar for many years as part of a deliberate government policy to ease export problems. The two dollars are now around parity with the increase in the price of oil but the Canadian did fall to as low 61 cents American. Canadian industry (not just manufacturers but primary resource producers as well) did not have to invest in innovation because the falling dollar would solve their problems. With the dollar rising with the rising oil prices, these companies are now finding themselves in a very difficult situation.

    Here in Quebec the forestry industry was formerly a major employer but has now all but disappeared. Ontario, the province with the largest economy, is having a very difficult time with its manufacturing centers. The province has entered into a green energy plan or boondoggle in response with dubious results. Ontario is being kept afloat by a financial services industry in Toronto.

    Canada does not appear to be a good example for either side in this debate. it has its own serious issues. Lack of investment in productivity improvements can certainly cost jobs. That was eh rs ult of deliberate government policy in Canada

  49. In regard to Canada and productivity, here is an article in the leading newspaper the Globe and Mail in which the Prime Minister of Ontario describes the harm that the high dollar has done to manufacturing. he does not indicate the the previous low dollar policy is the cause of the current difficulties.


  50. Hi Roger, I was studying for an economics exam and wanted to understand why my textbook said that increases in labor productivity will increase employment. In searching for the answer on the web I came across your article here. I'm a year late in responding, but I think I might be able to explain why you would actually expect to hire more people if productivity increased. It does require some principles of economics.

    Lets use the example of an industry where it takes an 8 hour shift for an employee to produce 1 unit of a good. The only cost to producing the good is labour costs.

    1. If productivity increases, you can create more output for each employee you have. Another way of saying this is that the cost per unit decreases.

    2. Lets assume that you will be able to easily sell any extra units you produce, without having to lower your asking price per unit. This is applicable if the market is large, say it consumes 10000 units, but you only produce 1 unit. In this case you have no bargaining power and you have to take whatever price is in the market. Also, any large % increases in your supply will not lower the market price. Or, a small change in price will cause a large change in quantity demanded - we call this elastic demand. We can assume other scenarios later.

    3. Per the concept of diminishing returns, your costs rise for each additional unit you create. Or you get less units for each additional dollar you spend. So let your first 8 hour shift for an employee cost $100. If you want that employee to work overtime and create another unit, it will cost you $150 for the second 8 hour shift.

    4. Let the price in the market for each unit be $120.

    5. You will produce 1 unit because it only costs you $100 but you can sell it for $120 and you make a profit. The second unit would cost you $150 which is more than the benefit of $120, so you will not produce a second unit.

    6. Now let productivity increase to 2 units per 8 hour shift.

    7. The first 8 hour shift costs $100. In that shift you can produce 2 units, or a benefit of $120 * 2 = $240. This is profitable, so you will continue to keep your employee for his 8 hour shift.

    8. The second 8 hour shift costs $150. In that shift you can produce 2 more units, or a benefit of $240. $240 > $150, so it is profitable to make your employee work the second shift too.

    9. Point 8 demonstrates how when productivity increases, demand for labour increases.

    10. Now lets revisit the assumption in point 2. Lets say the market is only willing to consume 1 units at the current market price of $120. The price must drop to $49 to encourage the market to consume 2 units, total spend of $98. We call this inelastic demand, a large change in price is required for a small change in quantity.

    11. After the productivity increase, the first 8 hour shift costs $100 and would produce 2 units. To sell both units, the price must drop to $49 each. This will yield a benefit of $98. (You can't charge $120 for the first unit, and $49 for the second. You can only have one price in the market) Since the cost of the 8 hour shift is $100, it is not profitable to produce 2 units. We could have our employee work only 4 hours, which would cost $50, and only produce 1 unit. In this scenario of inelastic demand we have cut employment.

    12. We have looked at 2 extreme examples, elastic demand which leads to an increase in employment, and inelastic demand which leads to a decrease in employment. Depending on the situation in your industry hopefully this can give some indication of what the end result will be.