*Innovation Economics*

by on September 23, 2012 at 7:49 am in Books, Data Source, Economics | Permalink

That is the new book by Rob Atkinson and Stephen Ezell.  It is far more mercantilist than I feel comfortable with, yet it is full of information and argumentation, and it is a book one can profitably engage with.  Here is one excerpt:

The largely consensus view among U.S. economic elites is that the massive U.S. job loss in manufacturing is simply a reflection of manufacturing doing well: using technology to automate work and to become more efficient.  It’s the agriculture story they tell us…

There are two big problems with this view.  The first is that it is not supported by the official government data.  In fact, U.S. manufacturing lost jobs much faster in the 2000s than in the 1990s, even though productivity growth was similar during the two decades.  In the 1990s, U.S. manufacturing employment fell 1 percent, while productivity increased 56 percent.  Yet, in the 2000s, manufacturing employment fell 32 percent while productivity increased only slightly faster, 61 percent.  So, clearly, higher productivity was not the main cause of the manufacturing employment collapse.

As Michael Mandel has pointed out repeatedly, there are also problems with the data, and here are our authors on that point:

…a closer look reveals that every durable goods industry grew more slowly in output than GDP except one: computer/electronics which grew a whopping 720 percent faster than GDP…To put this in perspective, this one sector accounted for 113 percent of U.S. manufacturing output growth in the 2000s, even though, in 1997, it accounted for just 12 percent of manufacturing output.

Note that a lot of this measured growth is quality improvement in computers, rather than growth of the sector in the traditional sense of having a rapidly expanding industry.  Employment in that sector fell.  The performance of the other manufacturing sectors is not so impressive:

…during 2001-2010, manufacturing minus computers actually lost 6 percent of its value-added.  Output of the electrical equipment and wood products industries declined by 7 percent, plastics by 8 percent, fabricated metals by 10 percent, printing by 12 percent, furniture by 19 percent, nonmetallic minerals and primary metals and paper by 31 percent, apparel by 34 percent, and textiles and motor vehicles by 39 percent.  In other words, thirteen manufacturing sector that made up 58 percent of U.S. manufacturing employment all produced less in 2010 than in 2001, all at a time when the overall economy grew 15.8 percent.

I suppose you could say that education and health care have in fact seen striking advances in productivity during this period.  Or you could recall my portrait in The Great Stagnation of an economy which has only a very small number of dynamic sectors, with computers of course in the lead.  Overall it is not a pretty picture.

1 stephen September 23, 2012 at 7:59 am

“In the 1990s, U.S. manufacturing employment fell 1 percent, while productivity increased 56 percent. Yet, in the 2000s, manufacturing employment fell 32 percent while productivity increased only slightly faster, 61 percent. So, clearly, higher productivity was not the main cause of the manufacturing employment collapse.”

Why assume linearity? Early innovation moves workers around, later innovation gets rid of them all together. Thats one possible model.

2 Bob Knaus September 23, 2012 at 8:20 am

Exactly. Plus, a consequence of improved manufacturing is that durable goods last longer, so fewer are needed and the consumer cost per year is lower. That, combined with a generally downward price trend globally for many manufactured goods, will tend to drive manufacuring “value-add” lower even though the consumer is receiving greater surplus.

Producers exist only to serve consumers. This needs to be said more often.

3 Tyler Cowen September 23, 2012 at 8:52 am

All fine and well, but keep in mind median real income during these years is also doing poorly.

4 Bob Knaus September 23, 2012 at 11:44 am

We are not as necessary as we thought we were.

5 Anthony September 23, 2012 at 12:55 pm

The simplest explanation is that there’s a lag between overall productivity improvements, and being able to reorganize one’s operation to use fewer workers. 10 years seems a bit much, so there’s probably something else going on, too.

I conjecture that median real income *overall* has done poorly, because most of those people no longer working in manufacturing are working in less-productive sectors, if they’re working at all. How’s the median real income *for manufacturing workers* doing?

6 Bill September 23, 2012 at 8:21 am

Just wait for low end service jobs get replaced by machines–ordering at a restaurant on your Ipad, going to a grocery store or retailer and simply walking through a checkout without a checkout person because all your items are chipped and all you need to do is show your electronic credit card with face recognition ID to pay.

7 Bill September 23, 2012 at 8:23 am

And, don’t forget, college profs being replaced with video online cartoon characters giving the lectures using the voices of famous actors.

8 Bender Bending Rodriguez September 23, 2012 at 7:54 pm

Will be able to chose? I’d love to hear Wallace Shawn talk about M&A.

I’m from Mattel. Well, I’m not really from Mattel.
I’m actually from a smaller company that was purchased in a leveraged buyout.

9 Brian Donohue September 23, 2012 at 8:24 pm

I dunno. The subtext, “what are we gonna do with all these people?” is a very old idea at this point. I think it undersells “these people”.

10 uffy September 23, 2012 at 11:48 pm


Yes, Luddites have been wrong for a good ~300 years at minimum, but I have yet to see any evidence that there are untapped opportunities in certain fields that can provide as many middle class jobs as has been the norm since WWII. The fact that the economies of almost every country regardless of median income are seeing these same problems – namely ever increasing debt required per new job – is not reassuring either.

11 Hürrem Hamam Otu Yağı September 23, 2012 at 9:24 am

Producers exist only to serve consumers. This needs to be said more often…

12 prior_approval September 23, 2012 at 10:01 am

I’m not sure that Turkish AI spambots aren’t the current bleeding edge of irony.

13 Jeffrey September 23, 2012 at 12:16 pm


14 Dismalist September 23, 2012 at 11:16 am

Relative employment in manufacturing has been declining in the advanced countries since ages ago. The cause has always been a combination of productivity growth and a deepening international division of labor.

Anybody want to stop any of this?

15 Neal September 23, 2012 at 11:36 am

How about price-factor equalization?

16 Keith Williams September 23, 2012 at 11:45 am

So much of our current economy is based on self-obsoleting electronica, rather than durable items such as houses and cars. Obviously, there are many problems with the overweighting of devices such as phones and tablets in our economy. Of greatest concern in the short term: there’s ~zero benefit to the US labor market. We all think Apple is leading us to greener pastures, but no- quite the opposite. Unless of course they start selling iHomes….

17 Keith Williams September 23, 2012 at 12:27 pm

… I should add, however that there is one very positive thing happening now in the American mobile sector: we’re already bumping up against hard bandwidth limits. With our population spread out far more than populations in Europe, we’ll need some heavy duty bandwidth infrastructure to support widely distributed consumer demand.

Much has been said about cash hoarding by companies like Apple and Google. There is a huge amount of potential investment (~trillions) that could hit the market in the form of partnerships to boost bandwidth (hence the recent return of Sprint stock). You can’t sell 5G phones and tablets unless there is a 5G network.

So, the smartphones themselves won’t really grow American jobs, but the need for bandwidth most certainly will.

18 JWatts September 24, 2012 at 4:48 pm

“You can’t sell 5G phones and tablets unless there is a 5G network.

So, the smartphones themselves won’t really grow American jobs, but the need for bandwidth most certainly will.”

No, not really. Or at least not enough to matter. Once the fiber is run, then you have to upgrade the switches every few years, but the cell towers are very durable. Once a crew builds and wires a tower, it’s good for a century. Sure, they’ll be some jobs, but not that many.

On the other hand, there will be a lot of jobs needed to generate all the content those devices consume.

19 Ray Lopez September 23, 2012 at 12:56 pm

The argument that there is a great stagnation in manufacturing are flawed, because it does not consider the rise of China. China is eating the USA’s lunch. So factor in China and there’s not a stagnation, just an outsourcing (to the detriment of the USA). That is my thesis.

20 derek September 23, 2012 at 1:21 pm

That is what I see in my industry. The large proportion of game changing innovation is coming from Asia. Asian companies are buying US based ones for their market and service assets from which they can sell. Almost all the asian manufacturers have products using the new technology, no US company does as far as I know. There is no stagnation. It is simply that the US is losing. Much of the discussion on manufacturing in the US is delusional.

21 prior_approval September 23, 2012 at 2:20 pm

‘Much of the discussion on manufacturing in the US is delusional.’
Shhh – the countries that support manufacturers prefer it that way. It is a very rare opportunity to exploit, watching your competition destroy itself due to its own illusions. You would almost think that some far sighted South Koreans (maybe the now dead convicted felon owner of the Washington Times?) or Germans (like the Bertelsmann Group?) would have been financing American media for their own purposes.

Naw – far too speculative. It isn’t as if extremely rich people finance information to influence the direction of politics or the marketplace, is it?

And as for real innovation – I still like mentioning Siemens manufacturing streetcars in California. First, they had to import welders, to teach Americans how to weld meter long sections of steel together, since they couldn’t find native ones to employ. Building streetcars from decades old obsolete plans, in part because the current lightweight composite and glass street cars with no step floors are simply unsuited for the U.S., whereas the old style hulking steel streetcars shrug off SUVs (actually, they pretty much destroy SUVs in a collision, proving the point that SUV drivers like to make that heavy steel makes for a safer vehicle when involved in an accident with a punier one).

In other words, when it comes to the sort of technology that Siemens routinely exports to a number of countries that Americans consider Third World, the American market is just too primitive to make it profitable. Much of the world’s active innovation is passing America by at this point – especially in areas like mobile communications, high speed data networks available to all, energy management systems, or a long line of automobile technology.

22 P. Ed September 23, 2012 at 2:28 pm

The Siemens anecdote is great–where might one find more details?

23 Zephyrus September 23, 2012 at 3:01 pm

Yes, the Siemens anecdote is really interesting. Don’t hoard your anecdata!

24 maguro September 23, 2012 at 3:11 pm

Yeah, streetcars are just the kind of cutting-edge 1890s technology that could lead us out of The Great Stagnation if only we could learn how to weld!

25 P. Ed September 23, 2012 at 3:21 pm

You don’t see much of a market for new transpo infrastructure?

26 Mark Thorson September 23, 2012 at 3:44 pm

We could make them driverless.

27 maguro September 23, 2012 at 4:16 pm

“We could make them driverless.”

Sure, but couldn’t you do the same thing with buses, without the disadvantage of being confined to a rail system?

28 Mark Thorson September 23, 2012 at 7:34 pm

Going driverless on rails should be orders-of-magnitude simpler than without rails. You don’t have to calculate your trajectory — you know exactly what your path will be all the time. The only variable is speed, so you reduce a two-dimensional problem to a one-dimensional problem.

29 David Bley September 23, 2012 at 4:42 pm

Electronics and computers are not industries that are contributing to job growth. Even if the product says “Made in the USA”, that product consists of modules and subassemblies made elsewhere and assembled and final tested in the US. So the US labor content is low. Sales of desktop computers, laptops, and consumer electronics are down – smartphones are up, but there is very little US labor content in them.

30 Bender Bending Rodriguez September 23, 2012 at 8:01 pm

Remind me again how little US labor content is in that Windows 7 license, that Office 2010 license, iOS, and Android.

31 Max Goldberg September 24, 2012 at 1:36 am

it is propably the best gold price chart on the market for investment http://www.sh1ny.com

Max Goldberg
http://www.sh1ny.com Founder.

32 Butalbi October 3, 2012 at 2:05 am

I am totally agree with your thoughts. Keep doing these type of work.

33 Bitalbutal October 23, 2012 at 6:47 am

there are many problems with the overweighting of devices such as phones and tablets in our economy.

Comments on this entry are closed.

Previous post:

Next post: