America’s industrial juggernaut

If you are as old as me, you might remember the 1950s and 1960s, when America was an industrial juggernaut.  Here are a couple pictures from Ford’s giant River Rouge complex, from (I’d guess) the early 1960s:

Screen Shot 2015-05-20 at 10.34.20 AMScreen Shot 2015-05-20 at 10.34.47 AM

Now America has de-industrialized.  And yet, we somehow managed to increase industrial production from 83.7 in June 2009 to 106.2 at the end of 2014.  That’s an increase of 22.5.  But 22.5 is just the change in an index number, what does it actually mean?  Well in March 1961, soon after JFK took office, America’s industrial production stood at 22.2.  That’s right, Obama presided over a boom in industrial production larger than the entire industrial production of the US in March 1961!  (And recall that IP is a real index, adjusted for price level changes.)

Misleading?  Off course, but still kind of interesting.  Note that America’s population has grown since 1961, but it hasn’t even come close to doubling.  Meanwhile IP is up nearly 5-fold.

PS.  IP peaked at just under 46 in November 1973, generally regarded as the date when the post-WWII industrial boom ended.

Do I believe these numbers?  Not really, as I don’t believe the government’s price level numbers.  Lots of this “growth” occurred in the 1990s and is just Moore’s Law in computers, not the US actually producing more “stuff.”  I don’t consider my current office PC to be 100 times better than my 1990 office PC.

 


Tags:

 
 
 

82 Responses to “America’s industrial juggernaut”

  1. Gravatar of E. Harding E. Harding
    26. May 2015 at 11:55

    http://research.stlouisfed.org/fred2/graph/?g=UBd

  2. Gravatar of Britonomist Britonomist
    26. May 2015 at 11:59

    “I don’t consider my current office PC to be 100 times better than my 1990 office PC.”

    You should. It can probably do thousands of more operations a second than your 90s PC.

  3. Gravatar of Randomize Randomize
    26. May 2015 at 12:05

    I’m with Britonomist. The things my PC can do compared to the old 386 (the one with the “turbo” button) are mind-blowing.

  4. Gravatar of Cameron Cameron
    26. May 2015 at 12:08

    “I don’t consider my current office PC to be 100 times better than my 1990 office PC.”

    Actually I think Moore’s law would imply your computer would be about 33 million times better than it was in 1990 (Assuming price performance doubles every year).

    That being said, I’m quite sure you personally get 100+ times more out of your computer than you did in 1990. You reach tens of thousands of people with your blog and have much more influence than you ever could have hoped without computer and internet innovations. A lot of positive externalities here too. In fairness, you’re an exception.

    I still think you’re right about happiness and utility being hard to measure.

  5. Gravatar of ssumner ssumner
    26. May 2015 at 13:00

    I have two main uses, word processing to type papers, and the Internet.

    The Internet is much better, I haven’t seen much improvement in word processing. It can do more, but is far more confusing to use.

    But of course that’s all subjective. I’d say my current computer is 3 or 4 times better, in terms of personal utils.

    Thanks for those graphs E. Harding

  6. Gravatar of TallDave TallDave
    26. May 2015 at 13:11

    And somehow air and water pollution are now much lower now too, and there are more trees in the U.S. in 2015 than there were in 1915. There are a lot of crazy numbers out there that aren’t widely known because people aren’t that interested in hearing how good they’ve got it.

  7. Gravatar of Jose Romeu Robazzi Jose Romeu Robazzi
    26. May 2015 at 13:20

    @Britonomist, @Sumner
    Both are right. Computers indeed can do even more than two magnitude orders of processing than 20 years ago. But that technological advance is not fully utilized either because: 1. much of that capacity is take up by the operating system, which is much more “robust” nowadays (less crashes, multithreading, better video, etc) than before; or 2. Users don’t really need all that computing power, and therefore only need a different device that should cost 100 bucks …

    But I have a different question, do you guys see any reason for durable goods / worker increase since 1990 and non-durable goods / worker be stagnant ?

  8. Gravatar of W. Peden W. Peden
    26. May 2015 at 13:28

    I recall PCs being MUCH more crash prone back in the 1990s. Apart from a period when I got a new laptop which couldn’t handle hardware acceleration, I have seen about two or three blue screens of death in the past 10 years.

  9. Gravatar of Matt Waters Matt Waters
    26. May 2015 at 13:33

    If I understand CPI quality adjustments, it assumes basically perfect markets for prices and substitutes. In other words, if 50 Pentiums cost the same as a Ford, then manufacturing 50 Pentiums is the same as manufacturing a Ford.

    It makes sense to me. If a completely new technology is introduced into whatever bundle you are looking at (industrial production or customer purchases), you have a base price in that year. Then, if a new computer is introduced and both the new and old computer are being sold at the same time, perfect markets will supposedly give the exact quality difference through the contemporaneous price difference.

    Even if computing power is “underutilized,” the underutilization will push down prices of computers. Therefore more computers won’t mean as much as new cars.

    The biggest issue is markets aren’t perfect. A new car lot could have both old model and new model cars at the same time, but the price difference probably doesn’t reflect quality difference. Maybe it’s close enough though. The fact that iPhone equals, say, a washing/drying machine in price means something, even though an iPhone doesn’t have the same tangible, big industrial qualities. (Really the Gorilla Glass is produced by America, as well as some other components. But the point stands.)

  10. Gravatar of Matt Waters Matt Waters
    26. May 2015 at 13:38

    Also, processing power is utilized IF the task is parallizable. Generally when people say OS’s haven’t kept up with processing power, they don’t realize memory size, hard drive access and network access speeds also need to keep up.

    Since GPU’s are built around processing and parallelization, I suppose you truly see the effects of processing speed improvements with computer games. And truly it is pretty stunning if you look at, say, the Unreal 3 engine. I played games in college 10 years ago and I figured graphics had improved, but I didn’t realize by how much.

  11. Gravatar of Anthony McNease Anthony McNease
    26. May 2015 at 13:39

    That’s weird. Shouldn’t the index just measure the amount of goods created/built in some year’s real dollars?

  12. Gravatar of Gordon Gordon
    26. May 2015 at 13:40

    “The Internet is much better…”

    Scott, by just lumping things under the term “Internet”, you may not realize how much you’ve gained under Moore’s Law. I recall that you participated in some sort of podcast from your home. Without the improvement in computing power to enable your computer to encode the video and audio, the only way you would have been able to participate in such a thing would be to travel to a studio with expensive equipment to handle the audio and video feeds. So many things that used to require travel and a physical presence have gone away. And that has far more to do with the increases in computing power than the Internet.

  13. Gravatar of W. Peden W. Peden
    26. May 2015 at 14:20

    For an interesting introduction to the connection between processing power and the internet-

    http://kernelmag.dailydot.com/issue-sections/features-issue-sections/12228/mac-plus-modern-web/

    We’ve become very used to a multimedia web. Even more importantly, we’ve become used to not having our eyes blinded by those awful geocities backgrounds and having your mouse changed to spew out sparkles.

  14. Gravatar of benjamin cole benjamin cole
    26. May 2015 at 15:14

    interesting post.

    In one way it doesn’t really matter what our industrial production is in the US, since we can print money and give it to offshore people and they give us back goods and services.
    That said, I also wonder if our industrial production numbers make any sense.

  15. Gravatar of bill bill
    26. May 2015 at 15:18

    Look at numbers for tons of steel and # of autos. You’ll be pleasantly surprised.

    My computer today is definitely 100 times better than what I had 20 years ago. It gets themoneyillusion.com! :-)

  16. Gravatar of Major_Freedom Major_Freedom
    26. May 2015 at 15:22

    Yeah, about that whole “I don’t trust the numbers” thing…

    Pretty much invalidates every claim ever made that the GDP and employment numbers in recent years is “evidence” that the recent rising NGDP is “working” as advertised.

    That incidentally has been my position the whole time. Can’t trust an institution shielded from market forces.

  17. Gravatar of Nick Nick
    26. May 2015 at 17:00

    A machine that only browses the web and does word processing now costs approximately $0, and is very light weight and portable. If that’s all you need, get that. You will have gained infinitely over your 1990s purchase, instead of merey X100 …
    I exaggerate slightly, but it’s really not the govt datas fault you bought more machine than you needed.

  18. Gravatar of E. Harding E. Harding
    26. May 2015 at 18:06

    I agree with Sumner on the computers. Abstract increases in processing power just haven’t been noticeable to me, an average user. I have a decade-old desktop and the only thing I don’t like about it is that it can’t run some newer programs and has a slower startup time. Gary North also said that for him, word processing hadn’t changed at all since the mid-1990s. Also, there has been a Lesser Stagnation in hard drive sizes since the 2011 Thailand floods.
    @Nick
    -$0? Cheapest I can find (and from which I’m typing this) is $60-$70. Maybe you can find something running Android at $20.
    “Users don’t really need all that computing power, and therefore only need a different device that should cost 100 bucks”
    -But how many gigabytes of free storage space do they need? 5 is too few, I’m guessing. In any case, I think this explains the popularity of Chromebooks and iPads (which I will never understand, and not just because they’re visibly overpriced).

    Smartphones, however, have, indeed, gotten cheaper and much higher-quality since 2008 (from when I first remember news about the iPhone). The cheapest good unlocked ones are under $80. You couldn’t say that in 2008!

  19. Gravatar of Britonomist Britonomist
    26. May 2015 at 18:42

    “Abstract increases in processing power just haven’t been noticeable to me, an average user. ”

    It’s not abstract though. Computers really are much more powerful and much more efficient than they were a decade ago (and exceptionally more since the nineties) in basically every sense. You’re simply not utilizing this extra power, but you would definitely notice it if you were involved or interested in graphically intensive software for instance.

  20. Gravatar of TallDave TallDave
    26. May 2015 at 19:37

    I’m surprised how many criticize Scott on the Moore’s Law thing. We programmers often see focus on technology over function, but Moore’s Law is really just an enabling trend — everything lives and dies by the application. If most of the utility of a computer for someone is in the word processer, the main benefit of today’s machines vs the 1980s for that person is that they can run a word processor much more cheaply (I’d guess for around 10% of the cost).

    Now, of course, for many other applications that is not true at all, many popular applications of technology (for consumers commercial and industrial) could not exist today, and there are future applications that don’t exist today.

    I do think it’s possible to underestimate how much these changes matter to other people— people forced to will choose smartphones over indoor plumbing, which is somewhat surprising to people who have never faced that choice — and indirect benefits abound, especially in industrial settings.

  21. Gravatar of dtoh dtoh
    26. May 2015 at 19:46

    Just a gut reaction, but looking at growth of GDP and at industrial production as a % of GDP, the growth in productivity implied by the Industrial Production Index seems about right when you take into account: higher efficiency of equipment, reduced labor, lower cost materials, less material usage, etc.

    BTW – The discussion about computers is interesting but it’s not a very big piece of U.S. industrial production.

  22. Gravatar of dtoh dtoh
    26. May 2015 at 20:10

    To put this in perspective, value add from lap top and desk top computer production is about 1/15th of the value add from making dog food in the U.S.

  23. Gravatar of benjamin cole benjamin cole
    26. May 2015 at 23:26

    Dtoh: Have you seen some the dog foods out there? Better than what they serve at some Michelin 5-stars…

  24. Gravatar of dtoh dtoh
    27. May 2015 at 00:11

    @bc
    You sound like a connoisseur!

    BTW – While we’re on the subject of dog facts, did you know that Tokyoites now have more pet dogs than children.

    If Abenomics doesn’t work, they’ll probably have to further downgrade to gerbils.

  25. Gravatar of Patrick R. Sullivan Patrick R. Sullivan
    27. May 2015 at 05:34

    ‘I’m quite sure you personally get 100+ times more out of your computer than you did in 1990.’

    In Russ Roberts’ EconTalk podcast with Hal Varian they both agree that the internet makes possible an education in economics that previously was only available to graduate students enrolled and on campus.

    Which made me chuckle, because Varian once took two weeks out of his life to debate me over errors in his book ‘Information Rules’. On the usenet site sci.econ. He never admitted that he (and his co-author Carl Shapiro)was wrong in his treatment of the issue in question, but I got the feeling he realized he was missing something–and David Friedman, who also participated, made no secret that he felt Varian was in error.

    That was around Thanksgiving 2002. Post hoc, propter hoc, Varian took advantage of that to gain fortune, if not any more fame, at Google. I’m not sure how to capture that advance in statistics, but clearly it’s a pretty big improvement.

  26. Gravatar of We are still here manufacturiing… | Industrial Insight We are still here manufacturiing… | Industrial Insight
    27. May 2015 at 06:06

    […] http://www.themoneyillusion.com/?p=29441 […]

  27. Gravatar of collin collin
    27. May 2015 at 06:09

    I don’t consider my current office PC to be 100 times better than my 1990 office PC.

    Isn’t the point you can buy a tablet that does 10x the work as 1990 desktop that costs about 10 times less? (And how often does modern crash compared to 1990 computer?) It ain’t perfect measurement but it sounds reasonably close in reality?

    In terms of secular stagnation, I do wonder how much of it is because companies are spending less on IT investment but getting more in return? I work in an office and the amount of time and effort it takes to rollout new programs is very small.

  28. Gravatar of guest guest
    27. May 2015 at 09:00

    E. Harding, a Raspberry Pi costs $35 or even less (well, it needs I/O devices, but these are also much cheaper), and it’s easily equivalent to an early-00s computer. And this is a newly manufactured product, not something used/refurbished. So 100x, or at least 50x, really is in the right ballpark.

  29. Gravatar of Vaidas Urba Vaidas Urba
    27. May 2015 at 09:28

    Scott, offtopic but very important:
    http://www.moneyandbanking.com/commentary/2015/5/26/do-central-banks-need-capital

    “the present value of the Fed’s future profits from currency alone is larger than its current $4½ trillion balance sheet, which is thought to be enormous”
    – the Fed could easily do much more QE!

  30. Gravatar of Justin Justin
    27. May 2015 at 10:05

    On the willingness to pay test, I consider modern computers much more than 100x better.

    If I had to choose between a brand new PC from 1990 for $1 or a $999 Macbook Air, I’d without hesitation buy the Macbook. I’d still choose the Macbook if it cost $3,000 and the 1990 PC was $0.03.

    Likewise with other electronics, such as TVs. I’d pay $1,000 for a 50in LED TV rather than pay $1 for a 36in CRT from the early 1990s without hesitation.

  31. Gravatar of Tommy Dorsett Tommy Dorsett
    27. May 2015 at 11:42

    Scott — Maybe this is part of the reason for the productivity slowdown: the capital / labor ratio hasn’t moved (on net) in 15 years. Most of that is likely due to the ngdp crash of 2008-2009. Bad supply side fiscal / regulatory policy probably explains the rest.

    http://research.stlouisfed.org/fred2/graph/?g=1cAn

  32. Gravatar of E. Harding E. Harding
    27. May 2015 at 12:00

    @guest, if you add a monitor to that (~$40), that still adds up to more than the listed price for the WinBook TW700:
    http://www.microcenter.com/product/439773/TW70CA17_Tablet_-_Black
    Again, “Maybe you can find something running Android at $20.”
    I agree pretty much all 1990s computers before about 1997 are generally terrible by modern standards. After 2003, however, they’re generally decent enough.
    @dtoh
    Semiconductors, computers, and communications equipment contributed enormously to Industrial Production of durable goods in the 1990s (over two percentage points by the late 1990s):
    http://www.federalreserve.gov/releases/G17/Current/g17.pdf

    -Age structure doesn’t matter. Mark Adomanis once wrote,

    Have you ever read an article in Foreign Affairs talking about the “Dying [German] Eagle?”

    @Britonomist
    Dollar-for-dollar, I consider computers today to be twice as good as those ten years ago (most of this is probably in the faster processors and operating systems today).

  33. Gravatar of Scott Sumner Scott Sumner
    27. May 2015 at 13:39

    Traveling today, so just a brief (iPad) answer.

    People seem confused about consumer valuation measurement. Economists use maximum willingness to pay. Suppose my 1990 PC cost $2000. That number doesn’t matter. What matters is maximum willingness to pay. Suppose that was $8000.

    That’s its value to me. Suppose $8000 is $5000 in 2015 dollars. Then for my current office PC to be 100 times better I’d have to be willing to pay $500,000 for it. Or my school would have to be willing to pay that much. It’s not worth that much. At that price point the entire economics department would be forced to share one PC.

  34. Gravatar of collin collin
    27. May 2015 at 14:07

    That’s its value to me. Suppose $8000 is $5000 in 2015 dollars. Then for my current office PC to be 100 times better I’d have to be willing to pay $500,000 for it.

    This sounds like the opposite of (from FT) the problem of Goldman Sachs claiming we underestimate inflation because we are underestimating the improvements in software. For example, Grand Theft Auto 5 is improvement over Grand Theft Auto 4 even though the games had the same price tag on them. (I know the example is a little eye rolling but across software there is probably truth.) So inflation and GDP is underestimated from software improvements are balanced by hardware improvements of 100x since the 1990.

  35. Gravatar of Gordon Gordon
    27. May 2015 at 15:06

    “Suppose my 1990 PC cost $2000. That number doesn’t matter. What matters is maximum willingness to pay. Suppose that was $8000.”

    Personally, I would not be willing to be spend more than $1 on a 1990 PC in today’s dollars. It would be lacking the interfaces I need to connect to the devices I have. It would not have built-in wifi. In fact, many consumer PC’s in 1990 didn’t even come with any sort of network hardware. (People forget how many peripherals had to be purchased for 90s PCs.) And even if I could find the necessary peripherals to hook it up to the internet, it would struggle to render web pages today. All of the changes to HTML and CSS require more computing power and the encryption key length requirements for HTTPS went up a few years ago.

  36. Gravatar of C8to C8to
    27. May 2015 at 19:08

    Never reason from a personal utility function.

    Society, or some research lab, would pay 100x for my current computer over my nineties one since it has 4000x to 16000x the memory (to pick a metric)

    Cell phones may have saved millions of lives, smart phones will save millions.

    Yes the word pro user doesn’t have much gain but that’s akin to saying I gain nothing from aero planes because I’m not a pilot.

  37. Gravatar of Morgan Warstler Morgan Warstler
    27. May 2015 at 23:09

    “I don’t consider my current office PC to be 100 times better than my 1990 office PC”

    That’s bc you are f*cking nuts.

    http://www.theguardian.com/technology/2015/may/28/jonathan-waldern-return-virtual-reality-as-big-an-opportunity-as-internet

    Scott, when tech guys say my RAM doubled last year, you don’t really think “WOW, those guys are getting 2x as much” – there’s no chance you’ll think Morgan with desktop 100Mbps is doing far more than with 10Mbps.

    And you’re wrong.

    But let’s not argue, imagine that INSTEAD you truly felt in your bones, “I consider my current office PC to be 100 times better than my 1990 office PC”

    Like you woke everyday and stared at your rig and thought HOLY SH*T! and the did things with it, that made what you did with it in 1990, feel like a telegraph.

    How does this change your analysis?

  38. Gravatar of Morgan Warstler Morgan Warstler
    27. May 2015 at 23:23

    “What matters is maximum willingness to pay.”

    LOL.

    Dude, half American millennials would give up having children to not be kicked off Internet forever.

    How much are your kids worth Scott?

    If our unit of measure is Darwin impulse to extend our genes, and the offer on table is : NO INTERNET FOR YOU FOREVER BY FIAT OF GOD’S HAND…

    Humans will give up babies.

    They’ll give up toilets, cars etc before the Internet too.

    The Internet is just behind WATER and AIR. Which is to say, not being dead does trump Internet.

    But even delectable food and it’s intense improvement on your lot in life – ppof, sacrificed to not be kicked off Internet.

    —-

    Scott, I think Water and Air are your best comparison.

    The only reason we price water at $100 a month is bc nobody charges us $5K (and doesn’t care if we die). Internet is valued based on what it takes a competitor to deliver it to you.

    “What matters is maximum willingness to pay.”

    punt on this one Scott, just say YEAH: Air > Water > Internet > Non-Soylent Food

    then tell me how to measure GDP again.

  39. Gravatar of Jose Romeu Robazzi Jose Romeu Robazzi
    28. May 2015 at 06:55

    @Scott
    Off topic, what you think of the idea that negative rates in EZ in fact imply a tighter monetary policy stance, since they may be incentivizing people to hold more cash ?

    https://mises.org/library/ecb-and-negative-interest-rate-game

  40. Gravatar of Njnnja Njnnja
    28. May 2015 at 07:22

    Then for my current office PC to be 100 times better I’d have to be willing to pay $500,000 for it. Or my school would have to be willing to pay that much. It’s not worth that much. At that price point the entire economics department would be forced to share one PC.

    Check out this blast from the past. Apparently in 1990 Bentley University (“Bentley College” at the time) invested a great deal to provide its students with access to on-line news services from outside of the library. The Pr1me minicomputer they mention was probably only $50K, but that is to handle only 150-175 hours per month (for the entire University!). So to handle the actual demand of users today would probably require many more machines (or one dedicated mainframe instead of a minicomputer), with a cost of at least $500K, and probably more.

  41. Gravatar of Patrick R. Sullivan Patrick R. Sullivan
    28. May 2015 at 07:37

    ‘Personally, I would not be willing to be spend more than $1 on a 1990 PC in today’s dollars.’

    In fact, it would have a negative value to you, probably.

  42. Gravatar of Benny Lava Benny Lava
    28. May 2015 at 11:06

    It seems like some people have a hard time grasping the difference between “I don’t get much use” and “can be used for”. And honestly Scott raises an interesting issue tangentially. For most people there is no reason to upgrade to a new computer. If all you do is shop online or watch videos then outside of malware and hardware failure a 5 year old PC runs fine and in no way needs replacement. And this is a problem for the PC industry as people upgrade less and less.

    Apple was smart to make the iphone difficult to change batteries which necessitates buying a new one every few years. If it wasn’t for batteries we might be seeing the same thing with phones.

  43. Gravatar of Andrew Andrew
    28. May 2015 at 12:51

    You just got a new proponent of NGDP Targeting at the Fed. Seems like you are making inroads.

    https://www.stlouisfed.org/~/media/Files/PDFs/Bullard/papers/Optimal-QE-2015May27.pdf

  44. Gravatar of Justin Justin
    28. May 2015 at 14:23

    Scott,

    Why is maximum willingness to pay then vs. now the correct metric to measure quality improvements? It seems to me like the right comparison is to see what people would be willing to pay for the older models if all historical product options were available along with the current model at the current price.

    Let’s just consider the evolution of the base model 13.3in Macbook Air, which was introduced in 2008, starting in the present and working backward (this takes out the element of fashion, as these models all look the same to the casual observer).

    The 13.3in 2015 Air can be had from Apple for $999. Now consider the 2014 Macbook Air. It’s almost the same as the 2015 machine, so it wouldn’t take much of a discount to persuade some buyers to buy it rather than the new model, perhaps $959 would be the market clearing price. In turn, there wasn’t a big jump from the 2013 to the 2014 model. Perhaps the 2013 would sell for $899. There was a significant improvement to the 2013 models vs. the 2012. The battery life jumped from 7 hours to 12 hours, and the base storage space from 64GB to 128GB. If sold today, perhaps the price of the 2012 Macbook Air would have to be as low as $649. The update from 2011 to 2012 was meaningful, with the base level RAM increasing from 2GB to 4GB and a moderate jump in processing power, and the camera was upgraded to 720p. Perhaps the 2011 machine would sell for $499. The 2011 upgrade changed the processors from Core 2 Due to the Core i5/i7 processors and added a backlit keyboard and Thunderbolt, so maybe the 2010 model would sell for $399. What about the base 2008 model? It had 80GB of storage but it was a hard drive, not SSD. It had a lower resolution display than the 2010 model, and a very slow Core 2 Duo processor. Maybe that machine would sell for $199.

    Isn’t this a reasonable way to gauge how much a particular product has improved?

  45. Gravatar of Ray Lopez Ray Lopez
    28. May 2015 at 15:20

    Sumner: “I don’t consider my current office PC to be 100 times better than my 1990 office PC.” And in the remarks section, it’s typical Sumner goal-post shifting. First he implies yesterday’s PC is not much worse than today’s (not true), then switches to a ‘well I could do nearly as much in yesterday’s PC, if you just count word processing’ argument, then, after being rebutted on both these points, switches to a fancy (and equally flawed) ‘consumer valuation measurement’ model.

    Pied Piper pipes on…surprised anybody including the dozen or so regular readers of this blog even care what Sumner thinks, but then again you could say that about Krugman and DeLong as well.

  46. Gravatar of dbeach dbeach
    28. May 2015 at 15:35

    I don’t understand this comment: “Not really, as I don’t believe the government’s price level numbers. Lots of this “growth” occurred in the 1990s and is just Moore’s Law in computers, not the US actually producing more “stuff.” I don’t consider my current office PC to be 100 times better than my 1990 office PC.”

    So far as I know the BLS does not use hedonic adjustments for PCs in its CPI calculation (they’re not listed here: http://www.bls.gov/cpi/cpihqaitem.htm), and anyway the US doesn’t produce many. *Phones* are probably a major contributor to US IP (thanks Apple!) but I don’t think they use hedonic adjustment for those either.

  47. Gravatar of Njnnja Njnnja
    28. May 2015 at 15:42

    @Justin:

    A great site with prices for old macs with specs and vintage is Mac of all trades

  48. Gravatar of Ray Lopez Ray Lopez
    28. May 2015 at 15:43

    @Andrew on the off-topic StLouisFed paper advocating a form of NGDP Level Targeting framework; the paper: “We consider a simple and stylized 241-period general equilibrium life cycle
    model of quarterly movements … One-period, privately-issued household debt and currency are the only two assets. We divide the population into two groups [credit users and cash users]”.

    Does this cartoon model sound like something you want to model a $16T+ economy on? Only a crackpot would advocate this. Far better to return to either a pure gold (coin) standard, a world currency like Nobelian Robert A. Mundel advocates, or some variant of the Chicago Plan (Google this), which is a non-fractional based reserve system. In short, we need more equity and less debt, contrary to the implications for society from another stylized and probably wrong model known as the Modigliani-Miller theorem.

  49. Gravatar of ssumner ssumner
    28. May 2015 at 19:15

    I don’t have time to answer all of the comments, but there is massive misunderstanding about “value.”

    Yes, 1990 computers are almost worthless today, but that has no bearing on the issue I’m discussing. The only relevant question is how highly would those computers be valued today if they were the only computers you could buy. And the answer is they’d be valued pretty highly.

    Here’s an analogy. I recently sold my two old tube TVs for negative $20 each. (I paid someone to haul them out of my house, as they were too heavy for me to lift.) But one was bought for $1300 in the 1990s, and the other was bought for $600 in 1984. And even these prices are not their actual value (including consumer surplus), which probably exceeded $10,000. When valuing something like a TV, the relevant question is the most you’d pay for it if it was the only TV brand available in the world. Let’s be honest, people enjoyed watching those old crappy TVs almost as much as the new ones. When a new one comes out it initially sells for much more, but is only purchased by the early adopters. Who is buying the 4K OLED TVs with 80 inch screens for $50,000? Only people like Bill Gates, for whom money doesn’t matter. Average people would pay more than for their current TV, but not that much more.

    Yes, the utility of PC has improved much more than TVs, but not 100 fold.

    I am pretty sure that IT was a big part of the 1990s IP surge, although I may be wrong. But don’t be fooled by the share of IT in IP, it’s very different from the share of GROWTH. Suppose IT was 10% of IP. Also suppose output of IT doubled each year, while the price fell in half. In that case nominal output would be stable (at only 10% of IP), but IT would be a massive driver of growth in IP.

    I don’t know if people are considering these points strongly enough.

  50. Gravatar of ssumner ssumner
    28. May 2015 at 19:16

    Ray, Do you have a link for Mundell saying Earth is an optimal currency zone?

  51. Gravatar of Philo Philo
    28. May 2015 at 19:31

    “When valuing something like a TV, the relevant question is the most you’d pay for it if it was the only TV brand available in the world.” Why not: “When valuing a 14″ TV, the relevant question is the most you’d pay for it if it was the only 14″ TV brand available in the world”? (But there are 15″, 22″, etc., TV brands.) Or: “When valuing an entertainment device (such as a TV), the relevant question is the most you’d pay for it if it was the only kind of entertainment device available in the world”? (No doubt there’s a technical term in economics for the point I’m making–‘scope dependency’, or some such.)

  52. Gravatar of Gordon Gordon
    28. May 2015 at 20:30

    “The only relevant question is how highly would those computers be valued today if they were the only computers you could buy.”

    If all we had were 1990 PC’s, I agree that they would be highly valued because all the different software and internet services would not exist. But then I don’t see how it’s possible to quantitatively evaluate the change in value for any information appliance as the evolution of such devices brings completely new products and services.

    If all we had were 1990 cell phones today, people would probably be willing to pay as much for them as they currently pay for smart phones.

  53. Gravatar of Morgan Warstler Morgan Warstler
    29. May 2015 at 03:15

    “The only relevant question is how highly would those computers be valued today if they were the only computers you could buy.”

    This is INCORRECT. Jesus Scott, no wonder you have been flailing around with this concept lo these many years.

    The value is done this way: Uber can’t be done on 1990. Nothing can. ALL FUTURE GROTWTH. Finding God is a platypus riding a comet -can’t be done without Internet / Smartphones etc.

    Scott, the problem is you aren’t confronting the argument I have made all along.

    Your great-grandkids will choose to get ina VR pod and never leave. They will consume less calories and electrons than kids 100 years before. They will consume NO physical atoms other than IV slurry.

    Now whether you like or agree with this future – forget it – instead ASSUME this is what happens, bc most assuredly, I promise you this is what we are building.

    How does you definition of value fit?

    How is a world completed created in VR, in which all living and non-living souls choose to spend their entire lives.

    How can we use you definition, which is:

    We wake up the brain in a pan (Scott’s great grandkid in a Matrix VR pod), and explain to it, it’s whole reality is “fake” and then show it Scott’s 1990 computer and tell it, it can only have a 1990 computer, and ask it how much it will pay?

    Scott, you are off your rocker.

    Value is: NOW THAT YOU HAvE A CAR, what is VALUE of a buggy whip, if you STILL have a car?

    It’s not “assume you can only use a buggy whip”

  54. Gravatar of Morgan Warstler Morgan Warstler
    29. May 2015 at 03:28

    Marginal value ceases to work in digital space.

    The only correct measures in digital are:

    Moore’s Law and Network effects.

    Computer power doubles and each new node (user) grows something near exponential value.

    Right now we have a split in economy:

    Atomic where marginal value will makes sense.

    Digital where marginal value has no meaningful purchase.

    ——

    Scott and other econos who like to mention Facebook, are not truly confronting the atomic / digital divide:

    1. Eventually there is no atomic consumption, so eventually the idea of marginal value is complete eradicated.

    2. You can’t use marginal value on the digital side of the economy – it’s a flaccid nothing concept.

  55. Gravatar of Morgan Warstler Morgan Warstler
    29. May 2015 at 03:36

    Example:

    I now use Slack instead of email with my team. It enables me too spend half as much time overseeing twice as much work – my advantage is that I use Slack, and other guys don’t yet. This is worth a ton – if I could pay to keep others from finding out about it, I would.

    But Slack will be used soon by everyone, and only charge $8 per user.

    There’s NO way to put a value on email based on assume there is no Slack.

    Just like there is no way to put a marginal value on email on 1990, when nobody had email. DO you put the credit on one new used added is easy, OR do you put the value on 1 new user makes it worth 10% more?

    The value of digital is what it makes POSSIBLE THAT GETS DONE, and could NOT get done without it.

  56. Gravatar of Ray Lopez Ray Lopez
    29. May 2015 at 04:16

    @Sumner – RAM on a world currency, see for example http://robertmundell.net/economic-policies/world-currency/

    @Warstler – I’m afraid Sumner is right. Though your comment about email not being measurable since it did not exist in the 1980s is right, economists do this calculation all the time via “substitutions” (the main phenomena behind what Sumner discusses). For example economists have calculated that refrigeration was not that radical as you might think due to the existence of iceboxes (I’ve seen this paper years ago). They also do this with so called PPP (Purchasing Power Parity) where they assume a meal at a fine TexMex restaurant like Chili’s in the USA is equivalent to a TexMex restaurant like Chili’s the Philippines (e.g., same name, same trademark, but NOT the same restaurant, in food, service or ambiance, trust me on this, the USA version is markedly superior). That’s why under a PPP standard China is the world’s second biggest economy but the reality is that in terms of quality they are far behind the USA (and even Japan in my opinion). But economists play these games all the time, they even have a “world unit of currency” that’s based on substitutions such as PPP.

  57. Gravatar of Njnnja Njnnja
    29. May 2015 at 04:28

    I think you are confusing “market value” with “value to me.” For a typical, household level powered computer, whatever the year, you are (approximately) the marginal buyer so how much you value it is very important. But to determine the worth of a computer that is so powerful that at time x, it is as powerful as the typical household level computer from time x + 25 years, you are no longer the marginal buyer and therefore how much you value it doesn’t impact the market price.

    Who is buying the 4K OLED TVs with 80 inch screens for $50,000? Only people like Bill Gates, for whom money doesn’t matter.
    Although I’m sure Bill Gates has a TV like that, the real buyer of those “TV’s” isn’t personal use; it’s retailers and convention centers and other entities who will use it as a giant public (or semi-public) display. They have budgets and there is a reasonably good market in those kind of large displays so if it costs $50K, then the market value is really $50K. As an aside, the largest “TV” in the world is in Cowboys stadium and cost $40 million.

    Going back to computers, you say:
    Lots of this “growth” occurred in the 1990s and is just Moore’s Law in computers, not the US actually producing more “stuff.” I don’t consider my current office PC to be 100 times better than my 1990 office PC.
    The difference between a 1990 computer and a 2015 computer is 100 times, not because the marginal buyer of a 2015 computer in 2015 gets a 100 times improvement from the buyer of a 1990 computer in 1990, but because the marginal buyer of a 2015 computer in 1990 thought that it was 100 times “better” (i.e., was willing to pay 100 times more) than for a 1990 computer in 1990. You can argue that it was wasteful for them to spend so much on it back then, and the extra utility they got out of it was frivolous or inconsequential, but you cannot deny the market price (what they spent on it).

  58. Gravatar of Dan W. Dan W.
    29. May 2015 at 05:21

    In 1989 I paid over $2000 for an PC with a 20MB hard drive and a 286 CPU. Today I can pay less than $500 for a laptop that is hundreds of times more powerful and far more versatile. I can do things with this laptop that were not even possible in 1990.

    Clearly technology has provided me a huge consumer surplus – for less money I am getting far more value. But in Scott’s defense here is the question that matters: How much would I be willing to pay for my laptop? Would I pay $10,000 for it? No I would not. Would I pay $5,000? Maybe. But then I would only have one such computer in my house and not the five I currently do.

    So my current PC may be 100s of times better than my 1990 PC but I do not value it that way. In fact in inflation adjusted dollars it seems I valued my 286 computer far more than I value my current laptop.

  59. Gravatar of Peter K. Peter K.
    29. May 2015 at 05:58

    Bullard comes out for NGDP path targeting:

    http://blogs.wsj.com/economics/2015/05/28/fed-should-consider-new-ways-to-counter-weak-growth-bullard-says/

  60. Gravatar of Morgan Warstler Morgan Warstler
    29. May 2015 at 06:17

    Ray,

    I laid it out. There’s no getting around it.

    TODAY ————————-> Ray’s great grand kids in VR pod.

    Whatever economic hedonic adjustments have been used in past with ATOMIC economy….

    It’s not applicable to every year from TODAY forward:

    1. Less calories (until it’s just slurry / Soylent).
    2. Less electrons (more powerful chips)
    3. Less atoms (until it’s just slurry).

    Ray / Scott,

    Along the way to your great grand kids in VR pods, we are headed towards 3D printing where basically every physical thing costs $1 per lb. We’ll hit that, and then nobody will even print shit anymore, the 3D model will just be beamed on their eyeball in their pod. NO ATOMS AT ALL.

    Now looks gents, everybody, you can poo-poo this but AT MINIMUM, this future I’m explaining, you should admit or at least KNOW – it’s the one that everybody doing tech – this is what we’re building TODAY.

    So you should have the stones to face the idea I’m presenting.

    IF we are headed to VR pods, and each day into that future, there is less atoms, less calories, and less electrons consumed…

    And people end up having INFINITE NON-SCARCE CONSUMPTION of EVERYTHING.

    A 80″ OLED tv? That’s nothing.

    A 360 degree 4K submersive experience wrapped everywhere you turn your head – costs LESS – it can paint the VR room with priceless artifacts, naked women, and still and room for a 80″ OLED TV – and it’s all on a screen 5″ wide – or a projector.

    Ultimately, nobody starts with this sentence:

    If Morgan is right, and my great -grandkids are going to be in a VR pod…

    Unless it’s a joke conclusion, bc there’s no way to measure GDP as we head from TODAY till then.

    NO WAY TO MEASURE IT.

    Why can’t people admit that?

  61. Gravatar of Justin Justin
    29. May 2015 at 07:49

    “Let’s be honest, people enjoyed watching those old crappy TVs almost as much as the new ones.”

    This is exactly my point, though. We’re trying to measure the quality of one product vs. another. Comparing the utility of a 1990s TV when people didn’t know there was something better to utility now for a 2015 HDTV, and I agree that there is little difference between a 2015 4K 55in HDTV and a 36in CRT Sony TV from several decades ago. Same would hold for the original Nintendo vs. Playstation 4, or a 2015 Ford Focus and a 1915 Ford Model T. And yet in each case, the modern product is far and away superior. A measure of quality that doesn’t reflect that isn’t a good measure of quality.

    I can’t accept the result that there is little quality difference between a Model T and a Focus because people need cars to get around and so would be willing to pay similar prices for both if they had to and there was only one option, and so most of the difference in price represents inflation.

  62. Gravatar of Justin Justin
    29. May 2015 at 08:12

    –“So my current PC may be 100s of times better than my 1990 PC but I do not value it that way. In fact in inflation adjusted dollars it seems I valued my 286 computer far more than I value my current laptop.”–

    I’m sure you do value your laptop 100s of times more than your 1990 PC. You’d only pay $5,000 for your laptop now, but what price would you pay for a 286? Even if it were the only option, would you even bother buying it?

    Perhaps this is just me, but if suddenly we could only buy 1990 PCs, I just wouldn’t own a PC, regardless of the price. The 1990 models are simply too handicapped relative to what I’ve experienced with modern computers for there to be a meaningful difference to me of owning a 1990 PC or not owning a computer at all.

    Likewise with television. If the only televisions available for sale were 1950s models, I just wouldn’t own a TV.

  63. Gravatar of Roger Roger
    29. May 2015 at 11:19

    Morgan is winning this argument, imo. You guys are not grappling with his arguments.

    I suspect economics is between transitions. It’s like the old physiocrats trying to grapple with the value produced by an industrial revolution. Their framework simply wasn’t up to understanding what was possible.

    Economics needs a fundamental restructuring to fully grapple with a world where most incremental value is being produced at pretty much zero marginal cost. I am sure there is a Nobel prize and a place in history for the economist who lights the way. Instead we watch the physiocrats argue between themselves.

  64. Gravatar of Benny Lava Benny Lava
    29. May 2015 at 15:15

    Another point to consider is the way the TV industry has failed to attract 4k buyers. The 4k tv is here and there are few buyers. Why? Well maybe it is the recession. Maybe it is the lack of 4k content. But maybe it is that consumers can’t see a big difference. 2k is good enough. Value and utility, people. Some people in this discussion can’t appreciate the market’s ability to put a dollar amount on value. Are we techno communists?

  65. Gravatar of Scott Sumner Scott Sumner
    29. May 2015 at 18:26

    Justin, You said:

    “You’d only pay $5,000 for your laptop now, but what price would you pay for a 286?”

    I’d pay $5000 for a 286, if that was the only type of computer available.

    It’s not clear to me if people don’t understand what consumer surplus means, or they don’t understand that when 286 PCs were for sale our current super powerful computers were not available.

    Roger, You said:

    “Economics needs a fundamental restructuring to fully grapple with a world where most incremental value is being produced at pretty much zero marginal cost.”

    Let’s see, I’ve probably done 20 posts talking about the new ZMC economy. Have you read any of them, or just decided I’m a physiocrat based on gut instinct?

    Yes, I’m sure there’s a Nobel Prize for the first real economist who figures out that lots of IT goods are produced at near zero marginal cost.

    Njnnja, You said:

    “because the marginal buyer of a 2015 computer in 1990 thought that it was 100 times “better””

    There were no 2015 PC computers in 1990, just big mainframes with that much power. And guess what, average consumers didn’t value those big mainframes enough to buy them, only corporations bought them.

    Gordon, You said:

    “If all we had were 1990 cell phones today, people would probably be willing to pay as much for them as they currently pay for smart phones.”

    Not quite as much, but yes, that’s basically my point.

  66. Gravatar of Morgan Warstler Morgan Warstler
    30. May 2015 at 01:03

    “Yes, I’m sure there’s a Nobel Prize for the first real economist who figures out that lots of IT goods are produced at near zero marginal cost.”

    But what if ALMOST ALL goods are produced at Zero Nominal Cost?

    We’d have to admit the economist who explained the REALITY of that thing, he’d get some dynamite money, right?

    ——

    Next “The Physiocrats!”

    “Received wisdom suggested that during a period of scarcity, a government should step in and forcibly lower the price of grain, so that people could afford to buy it. Governments might also choose to ban grain exports.

    “But Abeille argued that government intervention in the grain trade was self-defeating. With lower prices, he argued, grain producers would produce less. They would also make less profit””and therefore have less money to invest in the next year’s harvest. Government intervention, in other words, would disrupt the efficient working of the free market in grain, which would ultimately turn scarcity into a famine.

    So he argued that the government should step back and let local prices rise. Producers in other parts of the country would respond to the high prices, and the area would be flooded with grain. Problem solved. But Abeille recognised that there would be a time lag between the price rise and the demand response. And so according to some writers, the implication of Abeille’s argument is that people who cannot afford grain should be allowed to die. Efficient economic management trumped humanitarianism.”

    http://www.economist.com/blogs/freeexchange/2013/10/economic-history-0

    So right now we have this economy where:

    1. There is no famine.
    2. Nobody dies.
    3. There is zero marginal cost to produce most goods and services.
    4. There is zero time lag between the price rise and the demand response AND MOST ALL THE TIME prices just fall.
    5. Most people live provably better lives consistently YOY. Year after year after year, “wow I can’t believe it,” “whoduh thunk it?”

    And it’s not like I’m anti-NGDPLT, it makes TOTAL SENSE, but I have this competing Q nobody ever answers:

    If college debt goes away (and college campuses).
    If car loans go away.
    If commercial office space goes away.
    If Govt. buildings go away.
    If cost of public pensions goes away.
    If wars go away.
    And then if manufacturing goes away.

    Ok we have houses, and Amazon / Walmart, Data Centers, and some killer public spaces with dynamic local retail (thanks Uber for Welfare!)… it’s not VR and slurry yet, but let’s say this is the next 30 years…

    Where does all the capital go? The 5% return capital is supposed to come from WHERE? What real, non-digital, stuff needs THAT MUCH money?

    And why isn’t that 5% on the chopping block?

    It makes no sense.

    Is al the capital is expected to grow at 5% AND ALL be put into residential real estate?

    Because if so, I support a serious LVT, meaning bigger than we have ever modeled as acceptable before. Like GIANT LVT and progressive consumption tax.

    Income tax HAS to be out, bc the more the top 20% work, the more cheap free sh*t instanly delivered globally gets produced for the bottom 80%.

    Literally ANYTHING that reduces the hours put in by the top 20% from 100 hours a week to less – has an inverse exponential negative effect on the consumption of 9B people. An extra 1000 movies? Mindbending VR that replaces colleges, offices and business travel – bringing 9B up to US speeds?

    All created with human capital, virtually NO capital needed!

    Yep income tax has to be out.

    Ok, so after residential real estate, what’s left? Medicine? (i suspect piracy and globalism) Insurance? Against what catastrophe?

    Properly understanding the first in human history forces at work here, is definitely BIGGER as the both the physiocrats and what came next, our founding Econ fathers.

    It’s bigger, bc it goes on for all of human history, it’s like BC / AD stuff.

    Before Digital / After Digital

    It has always seemed like MM and out NGDPLT is supposed to describe the events above, and then show why it is specially prepared for this future everybody intuits, but can’t quite describe.

  67. Gravatar of Dan W. Dan W.
    30. May 2015 at 06:52

    @ Morgan,

    The marginal cost of digital replication is near zero. But humans can’t eat bits. The food, water and energy we need to enjoy the standard of living we have must come from scarce resources that entail real economic cost. Fantasies of a costless society are not new and they are never correct.

    That said your perspective demands answers. Technology is lowering marginal cost of production while at the same time taxes and regulations are increasing the marginal cost of labor. As such it is not surprising we are seeing a decrease in labor participation. What people decide to do besides work is nobody’s business. But if there are people who would like to work but the law says they are uneconomical at the government mandated price then there is a problem that policy makers should address.

    And there is the question of what is MM and what is NGDP if the value and price levels of economic activity are so nebulous.

  68. Gravatar of Ray Lopez Ray Lopez
    30. May 2015 at 07:14

    @Dan W. – fantasies of a costless society are not new, and in robotics the area of the curve where volume is low and costs are equally low (near zero) is known as the “Magic Kingdom”. Morgan Warstler apparently likes to sell concepts to investors, and he paints a good “big picture” as that is what excites investors. He’s good at what he does and optimism is a better sell than pessimism (which is reserved for the prospectus, written by lawyers). Who knows? Maybe this time is different and Warstler is right? We’ll be printing chocolate cake and disposable organs on our 3D printer soon…wait, they did that already.

    I want my virtual reality MTV with the naked chicks and with cyber-smell…for free, with an occasional non-intrusive advert in the upper corner of my field of vision. Warstler please work on making this vision into reality!

  69. Gravatar of Jim Glass Jim Glass
    30. May 2015 at 07:55

    Yes, 1990 computers are almost worthless today, but that has no bearing on the issue I’m discussing. The only relevant question is how highly would those computers be valued today if they were the only computers you could buy. And the answer is they’d be valued pretty highly.

    Why not consider what today’s computers would have been valued at then? The original commercial cell phone sold for $10,000 today’s money, market price, plus dollars per minute of use. And that was just for making phone calls.

    If a cell phone with today’s capacities — providing entertainment of all sorts, practical usefulness of all sorts like GPS, plus critical safety benefits too — appeared back then, clearly it would have had a market price of *well* over that $10,000. That’s what people would have been willing to pay. Why does this measure of consumer surplus disappear in our time just because we’ve accomplished the utopian achievement of creating such huge numbers of these phones that their massive supply has knocked their price down by much more than 95%?

    That seems like a massive write-off of a huge increase in consumer welfare.
    ….

    I have two main uses, word processing to type papers, and the Internet. The Internet is much better, I haven’t seen much improvement in word processing. It can do more, but is far more confusing to use. But of course that’s all subjective. I’d say my current computer is 3 or 4 times better, in terms of personal utils.

    My uninformed judgement is that your internet presence, on this blog and elsewhere, and within discussions of your ideas by others, has gained you a heck of a lot of professional credence, among people far more reputable than the likes of me. Perhaps it contributed to your recent relocation as well?

    If so, do you value all this benefit only on the same scale as being 2x or 3x the value of pre-Internet word processing?

  70. Gravatar of Jim Glass Jim Glass
    30. May 2015 at 07:57

    FWIW: In 1980 I bought a TRS-80 Model III with 48k memory, dual 32k disk drives (for truly ‘floppy’ 5-inch disks) and a 2.03 mhz processor, plus an equally expenseive daisy wheel ribbon printer, for $12,000+ of today’s dollars.

    It was one of the smartest buys I ever made in my life, as I was the first person at the company I was working at to computerize, it was a small but very successful and fast-growing entrepreneurial business, the owner really appreciated how I multiplied my productivity with it, and he shot me to the top of the firm as a result. In fact, I’m still collecting ample dividends on that to this day. It proved to be worth far more than $12,000 to me.

    Would I pay $12,000 for such a computing system today? Of course not (except perhaps as a collector’s item). My smart phone has literally thousands of times the computing capacity and I wouldn’t pay more than $600 for it. But that’s because of all the comparable cheap alternatives available.

    The first commercial cell phone sold in 1983 for near $10,000 in today’s money, plus dollars per call — and all it did was make phone calls. Imagine if back then today’s smart phone technology had became available … how much would the military have paid for it? Clearly many multiples of $10,000 per unit, for distribution to key personnel in vital functions. (As opposed to every high school kid working as a supermarket cashier — recently as I was in line to check out several of them were chatting comparing their models.)

    All these numbers show the huge amount of consumer surplus in these products today, which goes unrecorded in GDP figures and in all the wage and “welfare” arguments.

    Consumer surplus is the great unmeasured stat. And as the great Peter Drucker used to say, what is measured gets recognized, while what isn’t doesn’t, regardless of the relative importance of the two — the cause of many serious business management problems.

    Having profits and money income measured continuously in fine detail (and money consumption in somewhat lesser detail) while consumer surplus goes totally unmeasured creates a serious assymetry in our tallied-up economic data — which I submit in like way contributes significantly to the political-economic problems and disputes of our day (among economists near as much as any other group).

    The Nobel I’m waiting for will go to the person who figures out how to add consumer surplus to the national accounts in the same tables as GDP and income.

  71. Gravatar of Morgan Warstler Morgan Warstler
    30. May 2015 at 08:25

    Dan / Ray / Scott

    Greenspan has had a riff for years about the shrinking tonnage of GDP. It goes down every year.

    I don’t think it’s a sales job to say this trend never stops.

    Now look at this (posted yesterday on Youtube (200K views):

    https://youtu.be/f0CluQiwLRg

    Now it’s not exactly a true robot – but I’ at a point where I see something like this that just is jaw dropping every damn week.

    ——

    Now I get economics is a science. (Altho I’m not sure what Macro looks like if there is only one currency).

    And I get that economics has some basic laws, and I’m not going with any of those, but instead talking about the Internet like it’s 19 AD and tech Jesus has another 11 years before the world gets it.

    But if none of us can imagine a reverse in tonnage weight of GDP, that path is one consistent with 3D printing… and we’re about down to 1960’s level of electricity per capita… while bandwidth not just your pipe, but what your rig to your VR googles can do is just going BOOM!

    How do you not point to a dot in the future, one that as been very well hypothesized by countless Sci-Fi writers, where VR pods actually happens.

    OK, so now we have the dot.

    From TODAY to THEN, everything I described happens.

    Will we first have three generations of aluminum Ford F-150s, yes. Will they be driverless and shared by then, yes.

    Will there be college campuses for a long time? Yes, but experiencing ACTUAL campus at little or no cost in VR is going to be possible SOON.

    Mainly, I think Scott’s got a big enough brain to say something on this stuff that:

    1. tech guys agree with.
    2 Economists have not yet thought of or said.

  72. Gravatar of Benny Lava Benny Lava
    30. May 2015 at 09:35

    Jim Glass I am a little confused by what you mean about the consumer surplus going unrecorded. Better computers increase productivity which is recorded. I’m not seeing the missing consumer surplus.

  73. Gravatar of Vivian Darkbloom Vivian Darkbloom
    30. May 2015 at 10:48

    “Better computers increase productivity which is recorded. I’m not seeing the missing consumer surplus.”

    Benny,

    Perhaps you should think harder about the word “consumer”. It doesn’t mean “worker” or “business”. Think about it this way: VAT is ultimately borne by the “consumer”. Everyone prior in the chain (i.e., those out to earn a profit with the goods they buy and sell), get a refund of their input VAT. Those folks are not considered “consumers”, but “entrepreneurs”. I got a lot of consumer surplus writing this comment on my Macbook. I didn’t add one iota to GDP (nor did Apple get any credit for it). I did, though, enjoy a bit of “consumer surplus” with that little machine which is not recorded anywhere but in the comments section of The Money Illusion.

  74. Gravatar of Gordon Gordon
    30. May 2015 at 12:38

    Scott, I think I understand now. The change in computing power between the 1990 PC and the 2015 PC didn’t bring about increases in consumer surplus in orders of magnitude. What it did do is drop the marginal cost of delivering certain types of goods and service such that it is now possible to have a consumer surplus for those goods and services which was not possible before. Is that correct or am I still not seeing it?

    The analogy I have is that there is a 1990 key and a 2015 key. The 1990 key unlocks a limited set of doors behind which I can access goods and services that are of value to me. The 2015 key unlocks far more doors than the 1990 key. It’s not the value of the key that changed dramatically. It’s the amount of goods and services from which I can obtain a consumer surplus which increased dramatically.

  75. Gravatar of ssumner ssumner
    30. May 2015 at 17:57

    Jim, You said:

    “Would I pay $12,000 for such a computing system today? Of course not”

    I think you would, if it was the only computer available. Now suppose you can choose between two machines. One is the old one you describe, which produces $12,000 of value to you. Another is a modern one that is twice as good (and a million times more powerful) and costs $1000. In that case you’d buy the $1000 computer even if the old one (valued at $12,000) costs 1 cent. And that’s because the $1000 machine gives you $23,000 in consumer surplus, whereas the old one would give you only $11,999.99, even if you paid 1 cent. How much you’d pay is a very misleading measure of value, when an alternative sell for much less than maximum willingness to pay.

    Gordon, You said:

    “Scott, I think I understand now. The change in computing power between the 1990 PC and the 2015 PC didn’t bring about increases in consumer surplus in orders of magnitude. What it did do is drop the marginal cost of delivering certain types of goods and service such that it is now possible to have a consumer surplus for those goods and services which was not possible before. Is that correct or am I still not seeing it?”

    Yes, let’s work with this example. Suppose the 1000 fold boost in computer power enabled companies like Amazon. This company allows you to pay 25% less for the books and CDs you buy (yes, I still buy CDs). The gain in consumer surplus is better proxied by the 25% drop in goods prices, not the 1000-fold boost in “power.” (Yes, I know there are other advantages to Amazon; I’m quite willing to believe a much bigger than 25% cost reduction on average, but not 99%.

  76. Gravatar of Morgan Warstler Morgan Warstler
    31. May 2015 at 04:49

    “I think you would, if it was the only computer available. Now suppose you can choose between two machines. One is the old one you describe, which produces $12,000 of value to you. Another is a modern one that is twice as good (and a million times more powerful) and costs $1000. In that case you’d buy the $1000 computer even if the old one (valued at $12,000) costs 1 cent. And that’s because the $1000 machine gives you $23,000 in consumer surplus, whereas the old one would give you only $11,999.99, even if you paid 1 cent. How much you’d pay is a very misleading measure of value, when an alternative sell for much less than maximum willingness to pay.”

    Scott this is a just a shell game, and I never miss the pea:

    “Another is a modern one that is twice as good”

    Is wrong.

    it is 1MILLION times as good.

    And if you don’t think so, you shouldn’t do economics. And if you think “The gain in consumer surplus is better proxied by the 25% drop in goods prices, not the 1000-fold boost in “power.””

    Again, you shouldn’t be doing Economics. You are measuring atomic things and actually limiting what someone can receive to their income. And this is wrong.

    You buy CD’s. Scott, normal people have (note I didn’t say “own”) every song ever made. And the value of each song is determined by how much some dude name Scott Sumner still pays for his CDs. The fact that Scott, doesn’t simply go to Pirate Bay and get every song for free – well that’s Scott’s problem.

    Another mistake you make, you do not buy MUSIC, you are buying an Atomic piece of plastic – you OWN the plastic. I say this bc it’s very very important to future of Econ:

    You cannot own the digital. Anything that can be copied cannot be owned. Owned in our legal traditions handed down since Magna Carta has always reflected the notion of economic scarcity… there would be revolution in streets if food or oil or land could be copied and govt tried to maintain the idea of “owned” title.

    Since this is a TRUE FACT, we only apply the legal idea of owned, of govt title to atomic things.

    Even IP only really work in terms of “keep thief from getting a pile of money” – meaning MONEY can only be used as a claim on atoms or labor of another.

    I know all this is a lot to take in, but: turn the telescope around:

    1. It is impossible for govt to enforce no copy rules, bc everything digital (save one thing)can be copied It’s a law of physics.

    2. If there is no pile of money thief has amassed (Napster), there is no more claim on the atoms or labor of another.

    3. I’m of course right, see here: when people scream that printing money is immoral they do so, bc suddenly the govt is able to lay claim on atoms of labor – and you respond by saying “tough sh*t”

    4. Well guess what Scott? You don’t like this new model, where digital can’t be owned bc there is not even a pile of money to go seize? Well tough sh*t – we’re not screwing around here Scott, this is about hegemonic reality. Economics HAS TO BE DONE with the ruleset on the ground.

    5. The only digital thing that cannot be copied is Bitcoin, or some similar block chain tech, that’s WHY everybody is so gaga over it – bc a bitcoin address can only be attached to one account holder at a time. Note however: BTC while it has long way to go, can conceptually be understood to the end of govt as we know it. I’m just trying to get you to GROK that tech has the ability to MAKE IT IMPOSSIBLE for govt to have control of money supply -again, this is me raving like a techno utopian lunatic – I’m a Uber for Welfare Keynesian who wants to run NGDPLT to keep political order.

    But as someone who gets the bits, the point is BTC “ends” a problem that originally caused govt to exist – the need for a central authority to ensure title.

    SO the point Scott is: by getting rid of Scott’s ability to RESELL the digital copy of a song hat anyone can have for Free, Scott is left with a plastic disk. How much is a plastic disk worth Scott? Apparently you spend $10 on them – but most people think plastic is worth less than a penny – it’s trash the have to throw away.

    Are you getting $10 in value for that plastic disk Scott?

    No. bc guess what Scott? Value is not determined by your quiant atomic economics measures coming from the govt.

    The REALITY Scott, is that you are spending $10 on 10 songs, that you VALUE at $1 per song – and EVERYBODY else values them at $1 per song, they just choose to get into the free money airblowing machine and scoop up all the songs they want for free.

    If your measure of value doesn’t admit: either Scott is wasting $10 on a piece of plastic trash OR that everybody else is getting tens of thousands of dollars in free songs – you have a broken model Scott.

    Try using mine and see what happens!

    ——

    My life’s work started with a Tandy 80 – I spent all night making it play Silent Night for Xmas morning. Years later, long after BBS systems, and many surely now illegal activities, I trudged to local newspaper and sat in editor’s office to read the AP wire bc I had read already NYT and WSJ and every magazine at local libraries.

    When I got Mosaic on the fledgling Internet it was a jump from a wheel to an airplane.

    ALL the magazine newspapers, the AP wire these were basically a wheel and I had been VAULTED into the Wright brother’s plane.

    Since then, the plane has become a rocketship in terms measurable by your old “atomic” system.

    Scott, if I push you out 30 years and you have a 3D printer and a personal supply of recyclable / reclaimable commodities – copper, silicon, plastic, etc – and ALL IP – ALL 3D designs are free – you can own basically any physical object at push of button.

    You have what Scott? The marginal value of 50 lbs of raw materials?

    Are you nuts?

    No Scott, you do not get to say:

    “Another is a modern one that is twice as good”

    or this

    “The gain in consumer surplus is better proxied by the 25% drop in goods prices, not the 1000-fold boost in “power.””

    bc everything costs $1 per lb, BUT you ACTUALLY have EVERYTHING – just like you have EVERY SONG.

    Scott, you can’t keep trying to measure GDP based on an ever decreasing population of fools who spend $10 on a worthless plastic disk.

    You will simply become irrelevant.

    BUT, you can internalize the lesson, and I believe, you can explain how NGDPLT is the optimal policy NOW to the growing group of people who think GDP is a terrible measure.

    It’s because IMO, we can simply DISCOUNT the LT over time from 4% to 3% to 2% to 1% as more and more of the economy moves towards digital free.

    But that’s your job Scott, you’re the economist – just make sure you don’t tell everybody who now have every song ever that the value is in OWNING it, bc they can’t resell it, but the value is in HAVING if forever.

    And please, for the love of god, don’t admit in public that you spend $10 of trash.It freaks people out.

  77. Gravatar of Ray Lopez Ray Lopez
    31. May 2015 at 07:13

    Morgan Warstler on Scott Sumner: “You will simply become irrelevant.” – ! Warstler is a genius! He’s figured out Sumner to a T.

  78. Gravatar of Morgan Warstler Morgan Warstler
    31. May 2015 at 10:12

    Here is what is most important:

    https://twitter.com/pmarca/status/604973829245435905

    This conversation is one I have been having with you Scott publicly and privately for years.

    1. It WILL NOT GO AWAY. because
    2. We, the tech guys, ARE RIGHT.
    3. It ONLY GETS louder, until finally you Econos come up with a “OK, there has been no stagnation, and our new economic models will not require atomic growth or atomic measurement…. we will measure FREE and NETWORK effects more users increases value for everyone exponentially)”

    Ok. Great!

    Now then, NGDPLT is optimal policy bc it best deals with this once in a lifetime transition from an atomic economy to a digital one.

    “Here’s why….” – by Scott Sumner.

    Another great one:

    Let’s STOP PRETENDING THERE IS UNEMPLOYMENT – by Scott Sumner

    Hey let’s use this marvelous smartphone (digital deflation stuff) to put everyone on welfare to work, increasing consumption for the poor. You can even call it Uber for Welfare.

    And do a PAPER on it with Miles Kimball and Roger Farmer (other supporters).

    ——

    Somewhere in here, is a deep logic problem that Scott, I’m sure you can noodle out.

    FREE. Digital Deflation. Network effects. A once in human kind occurrence.

    The worst part would be if Central Banks mess up monetary, bc their models are not tuned to this future.

    Now why does a nominal LT protect us from this dangerous thing.

    Again, FORGET UNEMPLOYMENT, the danger is that Central Bank ruins economy by using old time atomic economic measures and keeping digital transition ordained by god to from happening as soon as possible.

    Make tech guys happy Scott and the Fed will adopt NGDPLT.

  79. Gravatar of TravisV TravisV
    31. May 2015 at 19:45

    David Glasner is brilliant, so these observations of his are forcing me to reconsider my priors:

    1. “My main point in that post was that the huge investment in building physical infrastructure during the years of urban renewal and highway building was associated with the mindless destruction of vastly more valuable intangible infrastructure: knowledge, expectations (in both the positive and normative senses of that term), webs of social relationships and hierarchies, authority structures and informal mechanisms of social control that held communities together.”

    2. “The challenge for any city is to find a reasonable balance between allowing individuals to organize their lives, and pursue their own interests, as they see fit, and providing an adequate supply of public services and amenities, while limiting the harmful effects that individuals living in close proximity inevitably have on each other. It is certainly fair to point out that unfettered market forces alone can’t produce good outcomes in dense urban environments, and understandable that Krugman, a leading opponent of free-market dogmatism, would say so, but he curiously misses an opportunity, two paragraphs down, to make an equally cogent point about the dangers of going too far in the other direction.”

    3. “It was also the interstate-highway system that enabled Walmart to achieve rapid growth in sales by locating its new stores very close to major interstates, which gave the new stores a huge advantage over their small local competitors off the interstate-highway system, competitors less accessible both to customers and to the large trucks delivering merchandise from wholesalers and manufacturers. Little by little the economic vitality of small communities off the interstate-highway system was drained out of them, causing many of those communities to wither and shrivel up.”

    For links, Google these blog posts at Uneasy Money:

    paul-krugman-on-tricky-urban-economics

    intangible-infrastructural-capital

  80. Gravatar of ssumner ssumner
    1. June 2015 at 12:08

    Morgan, You are not even addressing my argument. It makes no difference whether the gain is 2 fold or a million fold, people are simply misinterpreting the idea of consumer surplus.

    Look, you can say whatever you want. You can say a 4k TV is 4 times as good as an HDTV. After all, it’s got 4 times as many pixels. But it’s just empty words, that don’t connect up with anything in real life.

    I have perfectly rational reasons for buying CDs. The cost is so low compared to my income that they are virtually free. I’m donating money to artists I like. I avoid the hassle of learning how to use Pirate Bay–I have better things to do with my life, like tell Ray how dumb he is.

    TravisV, Yes, David’s great.

  81. Gravatar of Morgan Warstler Morgan Warstler
    1. June 2015 at 17:35

    “Look, you can say whatever you want. You can say a 4k TV is 4 times as good as an HDTV. After all, it’s got 4 times as many pixels. But it’s just empty words, that don’t connect up with anything in real life.”

    This is a throw away joke response to my argument.

    Scott,

    Assume I’m right – your great grandkids live their whole lives in a VR pod. We’ve moved passed 3D printing for everything with your $50 worth of commodities infinitely recycled, to nothing atomic is even created.

    I know this is incredibly hard for you to fathom – you still think artists make money on CD sales.

    But, just once, assume the tech guys are right, explain to me how GDP INCREASES without money printing.

    I know we can always print money.

    What I’m asking is how does Real DGP increase when the totality of human consumption is copyable digital / virtual items.

    It makes total sense in atomic space – scarcity exists. You buy a thing, it’s yours you can resell it. People need loans. People need to loan money.

    I can’t understand how to measure REAL GDP when the marginal cost of producing another copy of everything is zero.

    What’s more I can’t see how as economy goes from 60% atomic / 40% digital to 1% atomic / 99% digital – we keep finding capital returns. How do we ever grow RDGP, isn’t it all just inflation to hit NGDP target?

  82. Gravatar of ssumner ssumner
    2. June 2015 at 08:39

    Morgan, I’ve said from day one that any RGDP estimate is just a guess, nothing new there. NGDP is the only thing that can get measured.

Leave a Reply