30 January 2012

All Jobs are Service Jobs

I am trying to figure out (a) what a "manufacturing job" is, and (b) why many economists think that such jobs are in some way a special category of jobs.

My emerging view is that all jobs are service jobs and some such jobs involve the manipulation of tangible goods. In our economic accounting we classify some (but far from all) of those jobs that involve the manipulation of tangible goods (for instance, those that can be put into a shipping container) as manufacturing jobs, and others (such as in construction) as services. The distinction seems somewhat (entirely?) arbitrary to me and as apt to mislead as clarify our discussions of innovation and the economy.

Here is a specific example to discuss (Thanks AC!):
In this sprawling facility on Route 128, sporty Kia coupes and Volvo trucks are regularly taken apart and reassembled. Caterpillar tractors and Harley-Davidson motorcycles are put through exacting trials that test the latest advances in power steering and antilock brakes. Both Aston Martin Racing and the Penske Racing Team come here to shave seconds off their times.

But the 1,000-plus employees at PTC never touch a wrench or ball-peen hammer. Instead they develop and advance software that allows automakers to design, build, and service the latest automobiles rolling off production lines all over the world.

“The actual making of cars has moved to other parts of the world,’’ said Sin Min Yap, PTC’s vice president for automotive market strategy, “but the digital making of cars is thriving here.’’
Are the jobs at PTC "manufacturing jobs"? Are they "service jobs"? And, most importantly, why does such a distinction matter for our discussion of innovation, the economy and employment? (For initial background, here is how the North American Industry Classification System characterizes manufacturing.)

My view -- all jobs are service jobs. I will follow up on the consequences of such a view in subsequent posts.


  1. Dr. Pielke:

    Talking heads, pundits, and media types supporting current public policy [more accurately termed politico policy], are always in search of a statistical data point, holding forth statistics as fact, then making the leap that the data point constitutes complete evidence of an open and shut case…..have recently latched onto the data point of an increase in manufacturing jobs.

    One must realize that the mantra of “manufacturing job” is a panacea. That is, implicit to the “manufacturing jobs” mantra of the politico and their intelligentsia following, is that this particular job category, the rise and fall thereof, is the fix all. If manufacturing jobs are falling then this is the key to economic woes and the falling jobs number is due to some evil exogenous force. The rise in manufacturing jobs is a key to economic success and, of course, any rise thereof is directly due to politico policy.

    Obviously the panacea and the political wrangling around this particular panacea is political dupery. However, the greater political nitwitery is that manufacturing jobs, the decline thereof, is a world wide phenomena and not singularly a US phenomena. Manufacturing jobs on a world wide basis peaked years ago yet overall output continues to increase. Technology and automation, as in the agricultural sector of the past, are supplanting the need for human workers. This technology and automation supplanting the need for human workers results in better more innovative product at a lower real cost.

    Where the political dupery and political nitwitery merge is: if one is successful in attracting manufacturing jobs and builds upon the panacea, and if one looks forward considering that the current batch of dupes and nitwits being long gone twenty years hence, then one finds themselves in a 20 year future economy that has attracted an economic sector that’s worldwide employment opportunities have peaked years ago and which the unemployment is decreasing at an increasing rate while output increases at an increasing rate.

    How does one frame the panacea at this 20 year future point? That is to say, how does one explain to the people alive 20 years hence that dupery and nitwitery purposely attracted an economic sector that was in decline in relation to human employment? How does one explain that nitwits attracted a sector akin to 1920’s and 30’s agriculture declining employment, and now its 1950 and agriculture is shedding jobs at a maximum rate while output skyrockets. And when manufacturing runs its course and becomes the agriculture of today, employing 3% of the employed work force yet producing mass abundance, how does one frame the panacea?

  2. I don't think it's an unimportant distinction, even if it is fuzzy.
    But you might want to check out the three-sector hypothesis.

  3. Roger
    You should check this out they are similar www.swri.org/. There are quite a few others around the country. I've always been suspicious of these distinctions. I think the real distinction is what kind of good paying jobs are available to a HS grad or less. I guess that's un-PC.

  4. I'm wondering how useful an exercise this is. I mean, it sounds like meaning is going to be shrouded by a larger umbrella.

    For example, it could be said that "all water falling from clouds is precipitation", but there is no good in a meteorologist only saying that this day or that place will be receiving "precipitation". It is probably better to leave distinctions in place, or perhaps use more distinctions to make data even more useful and illuminating.

  5. Good question, especially given the current environment. My guess is that "manufacturing jobs" is shorthand for well paid jobs in industries dominated by unions. Such jobs can now be better described as "government jobs". On the other hand, manufacturing jobs may simply have emerged in contrast to agricultural jobs and the distinction had passing merit. From an macro-economist's perspective it is a difference without much meaning. The issues are more to do with the price/value of the output, its utilization of available labor skills and capital and whether it can be traded internationally.

  6. .

    Its all a bit nutty. IIRC if you work for a company with a manufacturing SIC code you have a manufacturing job. So if you are an in house accountant at GM you have a manufacturing job. If GM hires your outside firm to do some accounting work you have a service job.

    Perhaps I was misinformed. As they used to say when I was in the Navy, "Just because it is ridiculous doesn't mean its not right."


  7. Nope. You been going tot he Krugman school of economics lately? Simplify it. Consider you have one steam engine. A guy goes out and mines coal and now the engine can be run. Mining is a productive activity -- someome took something from a state where it was useless and brought it into a state where it is useful--this "made the pie bigger." This is not the same as a job where you, say, paint the steam engine a pretty color. that is an unproductive job, though in a complex enough society, there may be a host of reasons for painting it assorted colors, this activity would make the pie a little smaller if anything.

  8. Interesting, because I have long wondered what is meant by a "service job." But I suppose, a job provides a service, in theory, and therefore all jobs are service jobs? (and to what or whom ought we serve?)

    It kinda sounds like an standardized exam question.
    Service: Manufacturing::Job:_____

    And the right answer depends on the political interests of the exam writers.

  9. Roger -

    Do you see any non-arbitrary way to distinguish between someone who assembles jet engines for a living and someone who assembles Big Macs?

  10. When I think of manufacturing jobs I think of lower skill labor requirements. I think why some see a special distinction is that people with little education could often achieve a middle class lifestyle in one of these jobs.

  11. -9-Joshua

    "Do you see any non-arbitrary way to distinguish between someone who assembles jet engines for a living and someone who assembles Big Macs? "


  12. Speaking of "assemble"...

    Its been stated that “Buy American” really means “Buy Union Assembled” as many inputs are foreign made due to supply chains. That the assembly is merely promoted as “Buy American” and that promotion is mainly funded by union political funds. Hence the 90% non-unionized are influenced to buy union assembled i.e. assembled by the aristocracy of labor.

    Just saying…..

  13. “I am trying to figure out (a) what a "manufacturing job" is, and (b) why many economists think that such jobs are in some way a special category of jobs.” - Dr. Pielke

    Maybe its not economists that create the special category and its more a politicos creating the category.

    Maybe this helps as well, some insight from 1883:

    "The discussion of "the relations of labor and capital" has not hitherto been very fruitful. It has been confused by ambiguous definitions, and it has been based upon assumptions about the rights and duties of social classes which are, to say the least, open to serious question as regards their truth and justice. If, then, we correct and limit the definitions, and if we test the assumptions, we shall find out whether there is anything to discuss about the relations of "labor and capital," and, if anything, what it is.

    Let us first examine the terms.

    1. Labor means properly toil, irksome exertion, expenditure of productive energy.

    2. The term is used, secondly, by a figure of speech, and in a collective sense, to designate the body of persons who, having neither capital nor land, come into the industrial organization offering productive services in exchange for means of subsistence. These persons are united by community of interest into a group, or class, or interest, and, when interests come to be adjusted, the interests of this group will undoubtedly be limited by those of other groups.

    3. The term labor is used, thirdly, in a more restricted, very popular and current, but very ill-defined way, to designate a limited sub-group among those who live by contributing productive efforts to the work of society. Every one is a laborer who is not a person of leisure. Public men, or other workers, if any, who labor but receive no pay, might be excluded from the category, and we should immediately pass, by such a restriction, from a broad and philosophical to a technical definition of the labor class. But merchants, bankers, professional men, and all whose labor is, to an important degree, mental as well as manual, are excluded from this third use of the term labor. The result is, that the word is used, in a sense at once loosely popular and strictly technical, to designate a group of laborers who separate their interests from those of other laborers. Whether farmers are included under "labor" in this third sense or not I have not been able to determine. It seems that they are or are not, as the interest of the disputants may require.

    1. Capital is any product of labor which is used to assist production.

    2. This term also is used, by a figure of speech, and in a collective sense, for the persons who possess capital, and who come into the industrial organization to get their living by using capital for profit. To do this they need to exchange capital for productive services. These persons constitute an interest, group, or class, although they are not united by any such community of interest as laborers, and, in the adjustment of interests, the interests of the owners of capital must be limited by the interests of other groups.

    3. Capital, however, is also used in a vague and popular sense which it is hard to define. In general it is used, and in this sense, to mean employers of laborers, but it seems to be restricted to those who are employers on a large scale. It does not seem to include those who employ only domestic servants. Those also are excluded who own capital and lend it, but do not directly employ people to use it.

    It is evident that if we take for discussion "capital and labor," if each of the terms has three definitions, and if one definition of each is loose and doubtful, we have everything prepared for a discussion which shall be interminable and fruitless, which shall offer every attraction to undisciplined thinkers, and repel everybody else". - 1883, William Graham Sumner, What Social Classes Owe to Each Other.

  14. Spoken like an ivory-tower academic who never got his fingernails dirty in his life.

  15. Politicians and voters like "manufacturing" jobs, because they are code for well paid jobs that don't require much by way of qualifications.

    The issue is that people who do not get decent qualifications no longer have much chance of getting a decent paying job.

    The, entirely vain, hope is that bringing factories back to the West will somehow bring back decent jobs for uneducated Joes.

    Part of the swing of blue collar workers to the right in the West has been due to this. Instead of seeing the left as representing them, as it did historically, they hope that an industrial-friendly and protectionist-friendly right might somehow conjure good factory jobs.

    They are so committed to this, that they actually think labouring jobs makes them better people. Witness Mark B's entirely ad hominen comment above.

  16. Do you see any non-arbitrary way to distinguish between someone who assembles jet engines for a living and someone who assembles papers for publication in scientific journals?

  17. -11- Roger


    Do you also think there is no non-arbitrary distinction between durable goods and non-durable goods ("soft goods")?

    If you do think that distinction isn't completely non-arbitrary (although somewhat arbitrary - is being arbitrary a binary characteristic, can something be "somewhat arbitrary?") but also not meaningless, couldn't the quality of being manufactured act like a "moderator variable" on that distinction? In other words, the degree to which something is manufactured is related to its degree of "hardness" or "softness": an airplane engine is more durable than a Big Mac, so it has a stronger characteristic of being manufactured.

  18. Roger - I guess I should say acts "like" a moderator variable - as I realize that a moderator variable "moderates" the relationship between a dependent and independent variable (not a relationship like the difference between a durable and non-durable good).

  19. On hopefully a related tack, I recall the term visible and invisible exports. Invisible exports were generally earnings from services, particularly financial services including banking and insurance. Visible exports were agricultural, raw materials and manufactured goods.

  20. -7- Jeff,

    I agree with you to the extent that the act of fueling the locomotive is probably producing more wealth and benefit than painting it does, but not that painting the locomotive necessarily destroys wealth (making the pie smaller, in your words). In this example in particular, painting probably increases the lifespan of the locomotive by preventing rust.

    As you mention, there are probably other benefits, too, that may not be obvious or apparent. But it's also undoubtedly true that the operator spends a lot more money on coal than on the painting service.

    If the operator does spend more on painting than on coal (or simply too much) he's probably going to go out of business, not differently than simply paying too much for coal or any other factor of his production (e.g., featherbedding).

  21. Roger, yo may have a broad definition of what a service is but credit card companies have pretty narrow definition. When we first started accepting credit cards so that our large corporate clients could pay us more easily, we found some customers cards summarily rejected every bill submitted. It turns out that we classified ourselves as a "service" business (application of coatings on customer supplied parts) but to the corporate customer, service meant "personal services" which included escorts or massage parlors. Changing our classification "manufacturing" made the problem go away.

    By the way, anyone who thinks manufacturing jobs is for someone with few skills, think again. A lot of manufacturing requires a very high level of skill and training to do it properly. It's just that its often narrowly focused. Plus the accountability and competition can be brutal.

  22. What is a 'manufacturing job'.

    In traditional manufacturing there is a substantial capital investment in plant and equipment that would be 'costly' to idle. Hence a relatively low skilled work force can extract a premium wage from the owner of the facility by agreeing to forgo 'industrial action', I.E. Not idling the plant.

    In service jobs, there is little investment in 'plant and equipment'. Idling the plant thru 'industrial action' costs the owner only the profit he/she would make selling the service.

    Some critical service jobs also pay premium wages in exchange for providing a 'reliable service'. Examples are garbage collection.

    Having said all of that when 'politico's' are talking about 'manufacturing' what they are really talking about is jobs that will pay a premium for a given skill level that are not easily moved elsewhere.

    In the small new england town I grew up in we had a textile plant where all the spinning machines were driven by belts connected to a shaft connected to a water wheel driven by a river.

    The owners replaced the water wheeled driven equipment with electric motor driven equipment. The first 'industrial action' that occurred after the introduction of 'electricity driven' machinery caused the owners to permanently close the plant. It was simple enough to load the electric motors onto a truck and move on to greener pastures.

  23. Could it be that the creation certain jobs require more capital investment and are therefore perceived as providing more stable and long-term employment prospects? For example, it's harder to close down a factory than a software company.

    I'm not saying this is necessary right, just that it is perceived as being so.

    You could also point out that many service jobs require substantial capital investment in the form of education. But the nature of the capital investment by itself provides the basis of a distinction.

    For example, people would be more vulnerable to economic shifts in the service economy since the capital investment is more personal. If a factory closes down, the losses are born by the owner. However, if society suddenly needs less lawyers, a lot of people who personally put $150k or more into their education are going to suffer.

    This would provide a rational for more public investment in education and re-training programs in a service-dominated economy than in a manufacturing-dominated one.

  24. .

    "Part of the swing of blue collar workers to the right in the West has been due to this. Instead of seeing the left as representing them, as it did historically, they hope that an industrial-friendly and protectionist-friendly right might somehow conjure good factory jobs."

    Perhaps. It may be that as the core of the wealth producing class (makers) they recognize that the party of the takers is not their friend.

    The XL pipeline is a nice example where the blue collar worker needs political protection, but not protection from foreign sources.


  25. -16, 17 ...- Reiner, Joshua et al.

    I can come up with all sorts of categorization schemes based on various criteria, but the more general point here is to call attention to the purpose of categorization.

    As many of the commenters here note, categorization creates categories which then shape how we think, act, organize etc.

    We might debate whether or not there are essential characteristics of certain jobs that lend them to be in this or that category, but I would find that decidedly unpragmatic;-) A more important question would be what is the purpose of our categorization and what categories then serve that purpose?

    My opening answer to that question is that all jobs are service jobs.

  26. Dr. Pielke:

    "...all jobs are service jobs". Maybe: "all jobs demand a service skill of labor".

    Although wholly underappreciated, the agricultural revolution gave rise to the industrial revolution [see Harold Demsetz’s book From Economic Man to Economic System for a 177 page very insightful read] as collective farming ended and private property rights created a new farmer who had innovative ideas which required someone to design and build the agricultural innovation [bring the idea to fruition]. From these innovative ideas, to fruition of such ideas, being the dawn of the industrial age.

    The dawn of the industrial age began the wide spread use of the term manufacturing which basically was a reference to industrial production. Hence manufacturing jobs could be easily substituted for industrial jobs.

    Hence a manufacturing job is merely a job associated with an industrial setting. Stated alternatively, labor is labor, its merely the setting that differentiates sub classes of labor. Arbitrary labor categories such as agriculture, mining, manufacturing, etc. are clearly related to the setting the labor occurs within. However, labor is labor and in each case it’s the skill of labor which is demanded, albeit in different settings.

    One point not discussed is the output of industrial jobs. The output is either a sub component of final output e.g. bolts or final output e.g. dishwasher. The final output of industrial production is many times a labor saving device for the final consumer e.g. dishwasher. That the good produced gives the consumer more than one use [10,000 dish washes]. That is, the final output produces a service for the consumer into the future.

    The output of service jobs might be perceived as one and off. The service workers repairs the dish washer. The attorney prepares the will. The lawn maintenance crew mows the lawn. The service tech changes the oil. However, the value of the service worker and the value of the industrial worker, the labor skills thereof, is a determinate of demand and supply [pro football-service worker vs. steel worker-industrial worker].

    Maybe “comparisons” is worth a look. If one compares pro football players to the general labor employed by Alcoa one would say service workers are high paid. If one compares Target retail workers to Alcoa workers the industrial workers are high paid. That is, it’s the argument one wants to make among labor positions that creates the manufacturing jobs mantra not so much the true value comparison to the aggregate economy.

    As stated previously, world wide manufacturing employment peaked long ago [its not singularly a US phenomena]. Whereas in 1790 over 90% of the US labor force was employed in agriculture, less than 3% of the labor force is employed in agriculture today however output is exponentially larger. The same is the way of manufacturing.

    Maybe the best term, looking forward, for an industrial job/manufacturing job of the future is: robot.

  27. - 25 - Roger

    Somewhat related to W.E. Heasley's post - I've always thought that with a service job, the significant balance of value of the work lies in the quality of the relationship between individuals on the opposing sides of the financial transaction, whereas in manufacturing the significant balance of value of the work lies in a product that is manufactured.

    I used to see that as a non-arbitrary distinction in the sense that if a company is in a service business, it behooves them to prioritize the reality that it is the quality of the relationship between their employee and the customer that is paramount. It is a distinction that makes a difference (non-arbitrary) in terms of how a company runs a service business.

    Of course, that distinction breaks down when a worker is assembling a Big Mac. It is primarily the product that counts there - not the individual relationship between employee and customer. And when a jet engine is being manufactured, the "customer service" relationship between the producing company and the customer may be just as important as the quality/cost of the product. We could say that the competitive advantage of a Big Mac is the service aspect of being "fast food," but it is also the low cost of the product being assembled.

    So maybe you're right that the distinctions are arbitrary. If a traditional "service" company focuses too much on the quality/cost of the product over the customer/employee relationship it will suffer, but the same could be said if it focuses too much on the customer/employee relationship relative to the quality/cost of the product. Similarly, if a jet engine manufacturer focuses too much, relatively, on either quality/cost of product or the customer/employee relationship it will suffer also.

    I'm still not sure, however, that it's completely arbitrary to view the service industry as selling a "soft product" - in other words a customer experience - as opposed to a manufacturing company selling a "durable product."

  28. -27- Joshua,

    I think durability may be useful, but some services should probably be categorized as durable. The Big Mac wears out after a few hours, the hair cut after a few weeks. Painting a house may last for several years. A prepared will or other legal service may serve for many years.

    I don't think it's going to be easy to demarcate along customer relations vs quality and cost, either. In some markets, there's room for both (clothing from Walmart vs bespoke tailored suits).

  29. In regards to Joshua’s comment [27], he reminds of this aspect: many of the manufacturing jobs that have migrated overseas are very low end manufacturing jobs e.g. textiles. If those jobs became in-sourced [returned] many service sector jobs would pay as much as or more than these low end manufacturing jobs.

    Regardless of the job being categorized as manufacturing, mining, agricultural, service, etc. the flesh and blood human being doing the job is human capital. Human capital is added to some level of capital to create to create agriculture, mining, manufacturing or a service.

    Which then leads to this phenomena: All capital and all human capital is mobile. Most capital and most human capital, over time, migrates to the lowest environment of tax and regulation [regulation being an alternate form of tax].

  30. Roger
    I think we should not forget the difference between manufacturing or service JOBS and the different sectors of industry, which sometimes go by the same name.
    In the 1960s Daniell Bell, Peter Drucker and Fritz Machlup, among others, have claimed that the service sector will expand as a result of the expansion of the education sector. They were interested in this new, knowledge based service sector made up by a new class of knowledge workers.

  31. -30-Reiner

    Yes, agreed. Bell et al. were certainly on top of macro trends ahead of most. I'd prefer to keep the terms "knowledge worker" and "service sector" as distinct ... all jobs are service jobs, but not all workers are knowledge workers.


  32. Your classifications are not helpful. It would be far better to divide the workforce into:

    1. Those paid from taxes. (A drain on the nation).

    2. Those paid directly in the local currency. This includes much of what we call "service" today, and manufacturing for local consumption. (Broadly neutral effect on national wealth).

    3. Those paid directly or indirectly in a foreign currency. International financial services and exporting industry often fall in this category (Contributors to national wealth).

    OK, there's still room for a lot of debate, and there is going to be a lot of overlap between the three categories above, but these groupings are very useful in that they give some feel for the way the nation is progressing and future prospects for our standards of living.

  33. The industry classifications are set up to allow linking of NIPA, NAICS and Census Bureau-survey data which in theory allows one to track all industry and inter-industry transactions as well as labor employed. What is a "service job" and what is a "service industry" are different qustions. A lot of work went into reclassification of service industries for NIPA and NAICS/SIC during the past decade. You have to be very good to work your way through all of of this data including bridge tables with any understanding (... what qualifies one to become a "NIPA expert," "I-O expert," or "Census Bureau expert"??) The definition of service industry/company traditionally has been based on the final goods & services delivered. The comment "all jobs are service jobs" addresses the jobs question. Classification of service v. manufacturing industries is useful to provide insight on econ growth and development (transition, or transformation), e.g., using economic growth models with factor based technological / productivity improvement, understanding sources of productivity improvement (in the work by say Dale Jorgenson and Robt Stiroh, Robert Gordon, Edward Wolff, etc. All of which will lead to even better approaches in the future.