From what I can gather from several recent articles, many serious pundits have arguments with the views of ‘techno-optimists.’ A techno-optimist appears to be somebody who has blind faith in the power of technology to cure all ills, and particularly, to create economic growth. (See e.g., here, here, here, here, or here, and there are many more…)
I understand skepticism with an optimist, and asking someone naïve to ‘show me the money.’ As a rule of thumb, it is a good idea to be skeptical of bloggers and futurists who lack hard economic statistics.
That said, skepticism is an attitude, not a conclusion. Go ahead and ask for hard data, but do not close an open question. Sometimes innovative IT has had spectacular impact on the economy and sometimes it has not. It takes work to recognize the difference. Indeed, there is a set of norms and a range of standard tools for settling these questions – such as quality-adjusted price indices, revenue adjusted for inflation, and a range of economic accounting techniques.
This column (shamelessly) aims to educate journalists about some of the big open questions in research that studies productivity growth from innovative IT. While I am at it, I hope the topic informs graduate students about great topics for a thesis. Be forewarned, however. It is a little shocking how much (or how little) we really know about recent experience. Many of these questions are wickedly difficult.
The only thing I can promise with certainty is the absence of an easy answer.
1. Online pictures and videos are everywhere except in the productivity statistics. Tim Berners Lee invented the web to share graphs and pictures and other media, and nobody paid anything for licenses for the software. It spread everywhere, so we got a combination of low cost and wide use. Seems good for economic growth, right? To be sure, a space alien might see the web for the first time and might conclude that the web was invented primarily to share baby pictures and cat videos, and to illustrate men and women in, um, acts of procreation. Reflect for a thoughtful moment and really look beyond the superficial. There are millions of web pages. An enormous ecosystem of browsers and server software and networking infrastructure supports its use, and some of these actors – mainly Internet Service Providers – make some revenue, well north of fifty billion in wireline access. Online advertising is a little less. Is that it? How about online sales, which is less than eight percent of US sales revenues, but does not make large profits? What boost did the US get – a one-time surge in growth or a steady set of gains until now?
2. How much productivity does email produce? Email has a different history than the web, but its rise motivates almost the same types of questions. Nobody ever paid a licensing fee to make use of this invention. Last year over 200 billion emails were sent per day. To be sure, 75% of those emails are spam, but that still leaves a lot of quality communication. Where do we see those gains? Gmail and HoTMaiL, the two largest providers of email, charge nothing. Microsoft’s Outlook makes revenue, but that is not the contribution of email to GDP. Surely not. So how much does it contribute? When did email lead to growth – twenty years ago, when email largely displaced postal mail of letters? How about recently?
3. What were the gains from reduction in search costs? Who among you uses portals, search engines, recommendation sites, and online maps? OK, everyone can put down the hands. Remember how your parents used to fumble through the Yellow Pages? Remember how a cross-city trip involved laying out a great big paper map on the kitchen table? Remember when a cousin’s recommendation determined which restaurant in a new neighborhood you visited? Difficult as it might be for teenagers today to comprehend, everyone did manage in those dark ages. Still, today is much better. How much did the reduction in search costs bring to GDP, if anything, and when? Advertising supports some of this activity, which surely is only one aspect of a broad economic gain. Well, how do you measure those gains?
4. What were the gains from making the long tail available? Ebay, Amazon, Craigslist, and a gazillion sites made the long tail available – of books, music, clothing, memorabilia, and millions of other products. What an enormous variety lies at everyone’s finger tips. What does the long tail contribute to economic growth? There have been some estimates in specific product categories, such as books and shoes, but none across the entire economy. Were most of those gains realized in late 1990s? Do those gains today simply grow along with Amazon’s growth? Are most of the gains not realized as revenue, and, therefore, not measured?
5. Up to date online news is addictive. Is it productive too? The creation of online media ushered in an era for news-junkies. Again, hard for a child to comprehend how anybody checked the sports scores or learned timely financial news (without access to expensive Bloomberg terminals), but we did manage before. More deeply, what is the contribution of more timely information to economic productivity? Seems like many gains are not measured. If they are, where do those gains show up in national statistics? In which industries?
6. Did the rise of remote work change productivity? The cell phone and smart phone made work more mobile. The deployment of cloud is pushing in the same direction. There have been a few studies of a consumer’s willingness to pay for mobility, but very little about its effect on work. How did the economy gain from the introduction of mobility into work life? Just like the other questions: Some of the gains might show up in productivity statistics, or in prices, or in the hours worked, or in labor participation rates. Or not. How do we know?
7. How much did Wikipedia benefit the economy? Founded in 2001, Wikipedia is the third largest repository of human knowledge in the world today, exceeded only by the British Library and the Library of Congress. It is a top twenty site in every developed country. Traditional GDP measurement would value the advertising, but Wikipedia does not have any. Is Wikipedia’s value the displaced revenue at Encarta, Britannica, World Book and Colliers? Or is Wikipedia’s worth the value of the time put in by all its volunteers? Or is it the value to users, and the time savings it affords? What concept of value is appropriate, how would you implement it and measure its growth?
8. Enterprises do not own all their IT. Does that mean they are more productive? We are well past the days when an organization owned all its IT. Today we live in the era of cloud, networked services, rented manufacturing facilities, and marketing with social graphs. We are far past the era when a firm bought inputs and produced outputs in a manufacturing process and its productivity could be measured inside the manufacturing plant. In the extreme examples today firms such as Uber and Lyft own very few cars and employ very few people, and, yet, the firms are worth tens of billions. Has more value been created for GDP, if any, by these new organizational forms, that leverage external resources? How would you know?
9. How big were the gains from serving low density areas? Improvements in Internet access led to gains in GDP in places that had limited access to retail outlets. Traditional price indices have an urban bias, because the information is easier to collect in urban stores. Rural America comprises the experience of 15% of the US population (45m people), and a much higher percentage (and number of people) in the developing world. Again, how big were the unmeasured gains? Were most of the gains realized in 1995, or has there been steady progress since as the Internet has gotten faster?
10. What is the value of the creative commons license? A little less than twenty years ago some legal scholars invented the licenses known as the creative commons license. It is designed to permit sharing of copyrighted material in much the same way that open source has licenses for sharing code. Millions of web sites operate with this clever legal adaptation, it is integral to some popular sites with user-provided content, such as YouTube. What is the value of this invention? Do the price indices properly capture the extraordinary decline in the cost of sharing photography and video recordings? If not, how would you make a more accurate index?
These questions presume that many economic gains from the deployment of the commercial Internet went largely unmeasured. Blame flaws in traditional price indices, inadequate definitions of revenue, and flawed productivity attribution exercises.
These questions also tend to imply that the processes that created growth in the recent past will continue to create growth in the future. That motivates the research question. While it is good to fix recent growth statistics, it is even better to measure the gains in the near future.
If these two features make me an optimist, then call me that.
More to the point, if you’re an economic researcher and you do not like this list of ten questions, then make your own. Let’s focus on one of the biggest unaddressed economic topics of our time.
Copyright held by IEEE. To view the printed essay, click here. (edit)
11 Replies to “Ten Open Questions for the Techno-Optimist”
11. Where is the value in my being able to do my own plumbing/carpentry/landscaping/etc in less time than it would have taken me to call/wait/clean-up-after/redo-myself/pay-bill-from based on Youtube videos and parts I never knew about but can now find, order online, and have delivered to my door? Less plumbing/carpentry/landscaping/etc in GDP, but more leisure time, better results (look at **this** for example), increase in personal wealth, big boost in the never-measured, never-mentioned (after chapter 1) “personal utility”.
Good point! The last twenty years has altered the boundary between revenue-creating activity and so-called leisure activity that does not create revenue. Big broad question. Certainly there is plenty of evidence of changes in the use of time (e.g., using the Internet instead of reading newspapers or magazines), and there is plenty of room to dig deeper. Thanks.
Let us not forget that the contribution of the labor of homemakers to the economy has been largely ignored for decades
«do my own plumbing/carpentry/landscaping/etc in less time than it would have taken me to call/wait/clean-up-after/redo-myself/pay-bill-from based on Youtube videos and parts I never knew about but can now find, order online, and have delivered to my door?»
That’s called “consumer surplus” and it is similar to the value to you of using a well paved road to go do your shopping at a mall.
But GDP does not account for “consumer surplus” but for the gross value of domestic production. Attempts to add to it “consumer surplus” seem to me transparently aimed at adding imponderables to make it look bigger.
Not that such attempts are anything new: in the USA GDP is “methodologized up” in various ways, for example “hedonics”, or “quality” adjustements:
«Likewise, quality improvements – better cars rather than just more cars – account for much of the increase in GDP nowadays. But assessing quality improvements is difficult. Health care exemplifies this problem: Much of medicine is publicly provided, and much of the advances are in quality.»
«The most current figure I have for hedonic adjustment to the GDP is 2.257 TRILLION dollars which is roughly 22% of the GDP.»
Much of recent reported GDP growth is actually “methodological improvements”
The questions above seems exceptionally naive to me, because productivity is measured in terms of *final* outputs, and essentially all the questions above are about *intermediate* outputs instead.
If IT has improved productivity this will necessarily show up in form of improved productivity in final outputs.
Car companies don’t buy IT equipment because they just like IT, they buy IT because it helps to make cars. If IT increases productivity then that shows up in more car being built with the same budget, or same number of cars built with a smaller budget. For cost analysis it may be interesting to isolate the contribution of IT, but for productivity analysis all that matters is value added to the final output, the car.
In this IT is no different from the production of bolts and nuts or roof tiles: the vast majority are used as intermediates for producing something else, so changes in the productivity in their making are reflected in the productivity of making the final goods.
For example Google buys a lot of IT, not just because they love it (they do!) but because they use it to run an advertising business. If their IT improves in productivity that will be reflected in the productivity of their advertising business.
There also much less common cases where IT is a final output, in the sense of consumer purchases, or building up capital, but essentially all of these can be classified as “entertainment” (email, youtube, Netflix, …), with modest exceptions.
And these can be largely accounted for by the purchases of IT equipment by consumers to enjoy that final consumption, plus subscription fees. The final output of Netflix for example is hours of consumer entertainment, and the value of IT is reflected in their monthly fee plus a share of the cost of the PCs people buy.
I also think that a small part can be accounted for as building capital, and things like Wikipedia, The Internet Archive, and in part Youtube, Flickr, Blogspot, can be considered capital building, but that does not seem huge to me, even if perhaps the long term consumer surplus from that capital building will turn out to be huge, a bit like investing in highways was in the past.
That IT is mostly intermediate “goods” is blindingly obvious, so it is hard to understand why these questions get asked, unless the goal is to “methodologize up” the reported GDP index by double counting some intermediates as final outputs, which seems to have become a popular “methodology” with governments desperate to “improve” the debt/GDP ratio.
For the value of Wikipedia, Google Maps, etc. Is hedoniics the answer? How much would I have to pay you to give them up?
But is not about productivity, but maybe consumer surplus.
Consider Wikipedia, OpenStreetMap, GNU/Linux, for the sake of argument,as they are an extreme case all of them being in effect charities, non profit volunteer based projects.
They are not that different from government or local government projects, like a new library, or a new bridge, or roadsigns that indicate which way to go.
All of those are obviously “valuable” in some sense. People pay local and national taxes to get them. Probably the taxes paid are well below the consumer surplus they bring. J Stiglitz in the same article writes as to the difficulties of accounting for government spending in GDP:
«For example, while GDP is supposed to measure the value of output of goods and services, in one key sector – government – we typically have no way of doing it, so we often measure the output simply by the inputs. If government spends more – even if inefficiently – output goes up. In the last 60 years, the share of government output in GDP has increased from 21.4 percent to 38.6 percent in the U.S., from 27.6 percent to 52.7 percent in France, from 34.2 percent to 47.6 percent in the United Kingdom, and from 30.4 percent to 44 percent in Germany. So what was a relatively minor problem has now become a major one. [ … ] But assessing quality improvements is difficult. Health care exemplifies this problem: Much of medicine is publicly provided, and much of the advances are in quality.»
But as to productivity, they all either increase *final* output, in which case their productivity show up as greater output with the same inputs, or they don’t, and then they don’t increase productivity.
“Hedonics” is a bit of trick, the claim that somehow a trip by car is more productive in a nicer car than in a crummy car even if the duration and cost of the trip are the same. But while it may be more enjoyable that is not at all the same as more productive.
It is bad enough that most people confuse GDP, a vector of physical quantities, with the GDP index, which is the same multiplied by a vector of prices.
Going back to “How much would I have to pay you to give them up” consider that question applied to a collection of Monet paintings or to the children of a father. It might be a very large sum, but that sum does not measure Monet’s productivity or that of the father and mother, but rather the happiness of the owners of the paintings or the father of the children, an emotional, metaphysical, quantity and thus not properly part of discussion of economics and productivity.
For an extreme example consider heroin: it brings so much psychological relief to some tormented lives that users are willing to pay extraordinary prices to keep their habit. Does that mean it is extraordinarily more productive than aspirin?
If any of these things had significantly improved productivity, it would show up in GDP growth. GDP growth has been very slow precisely during the same time frame that information technology has been making this allegedly world-changing advances.
It may be that IT has improved people well-being in ways not measured by GDP, but that’s a very different question.
Non-economist follow-up question: Does the calculation of productivity include the income lost due to the internet- and IT-powered factors you list? Case in point, close to home for me: the magazine publishing industry did very well for decades, providing information, recommendations, and advice. It had two revenue streams (subscriptions and advertising). Now this once-powerful industry is a withering shell of its former self, thanks in part to several factors on your list.