Digital technologies aren\’t just changing the way existing companies communicate and keep records, but are creating new kinds of companies (think Uber, AirBnB, or Amazon) and products (think and \”free\” products like email and websearch or an app like Pokemon Go). Can the old-style methods of measuring GDP keep up? Nadim Ahmad and Paul Schreyer of the OECD tackle this question in \”Are GDP and Productivity Measures Up to the Challenges of the Digital Economy?\” which appears in teh Spring 2016 issue of International Productivity Monitor, which in turn is published by the Ontario-based Centre for the Study of Living Standards. Perhaps a little surprisingly, their overall message is upbeat. Here\’s the abstract:
\”Recent years have seen a rapid emergence of disruptive technologies with new forms of
intermediation, service provision and consumption, with digitalization being a common
characteristic. These include new platforms that facilitate peer-to-peer transactions, such
as AirBnB and Uber, new activities such as crowd sourcing, a growing category of the
‘occasional self-employed’ and prevalence of ‘free’ media services, funded by advertising and ‘Big data’. Against a backdrop of slowing rates of measured productivity growth, this has raised questions about the conceptual basis of GDP, and whether current compilation methods are adequate. This article frames the discussion under an umbrella of the Digitalized Economy, covering also statistical challenges where digitalization is a
complicating feature such as the measurement of international transactions and knowledgebased assets. It delineates between conceptual and compilation issues and highlights areas where further investigations are merited. The overall conclusion is that, on balance, the accounting framework for GDP looks to be up to the challenges posed by digitalization. Many practical measurement issues remain, however, in particular concerning price changes and where digitalization meets internationalization.\”
The article employs a refreshingly down-to-earth strategy: it discusses, one by one, certain kinds of transactions in the digital economy, how the digital economy has altered (or in some cases created) these transactions, and how well they are captured in GDP.
For example, one set of digital economy activities is what the authors cal \”intermediation of peer-to-peer services,\” which is hooking buyers up to sellers through Uber, AirBnB, eBay, new ways of getting loans, and others. The quantity and value of these kinds of web-based transactions has surely risen. But by and large, the value of these transactions are captured pretty well through the recordss of the companies involved. In these areas, one could argue that these underlying economic activities might be better captured as part of the digital economy than it was before. In the past, activities like unlicensed or off-the-meter cab drivers, informal off-the-books rentals, and garage sales were not well-captured in official economic statistics.
Sure, some tricky issues do arise here. For example, if I use my car as an Uber driver, then my car is no longer solely in the economic category of \”durable goods consumption,\” and now is also in part a form of \”business investment.\” But it also true that people who work from home, in one form or another, have been mixing the \”consumption\” and \”investment\” categories for quite some time now.
A different set of issues arises thinking about how the digital economy has enabled consumers to take over certain tasks previously provided by producers. Here\’s their explanation:
Perhaps the best example is the use of internet search engines or travel websites to book flights and holidays, previously the preserve of a dedicated travel agent. But there are many other examples that merit consideration under this broad umbrella where market production blurs with non-market activity: self-check in at airports, self-service at supermarkets, cash withdrawal machines and on-line banking to name but a few. These innovations have all helped to transform the way consumers engage with businesses and brought with them associated benefits but they also involve greater participation on the part of consumers, and indeed involvement in activities that used to be part of the production process. Because the involvement of the consumer displaces traditional activity, the question is whether this increased ‘displacing’ participation should be included in GDP, one of the main arguments being that GDP would be higher, for
example, when a travel agent acts as an intermediary to conduct the search compared to when the individual conducts the search his/herself.
But of course, this issue isn\’t new either. GDP has always been about what is actually bought and sold in the economy, not about what might have been bought and sold. There are lots of goods and services for which households have some degree of choice between making or buying: cooking, cleaning, child-care, assembly (say, of new furniture), home maintenance or decorating, transportation various leisure activities, and others. The authors argue that in this broader context, \”the scale of ‘digitalized’ participation activities is likely to be significantly less than those for other non-market services outside the production boundary.\” The usual approach to these activities for government statisticians is to set up \”satellite\” accounts in addition to GDP that offer estimates of their value, without actually adding them to GDP.
Some of the hardest issues arise in the areas of digitally-based consumer products that are free or subsidized to the consumer, like email, web-search, computer storage space, free software for computers, free apps for smartphones and tablets, and much more. Ahmad and Schreyer point out that \”it is important to note that the provision of free services by corporations to households is not a new phenomenon. Households have long been accustomed, for example, to receiving free media services (television and radio) financed implicitly via advertising.\” Historically, what you pay for a daily newspaper has mostly covered the delivery costs, while the cost of news-gathering and production was supported by advertising. In addition, it has been a fairly standard marketing approach in the past to give away a good or service at a free or reduced price, and in that way to try to encourage buyers to spend more afterwards.
Of course, some puzzles arise here for GDP statisticians. For example, one view is that \”the
value of the free service provided to the consumer can be equated with the value of the corresponding
advertising services.\” Another view considers \”the time spent by households watching advertisements as an act of production, for which they are paid by the advertising firm, and in turn pay for the (previously free) services to the service provider.\” Various complexities arise here, but the differences in thinking about advertising-supported services are not fundamental in nature.
However, greater complications arise when part of the tradeoff for \”free\” digital services involves information. As the authors point out, the advertising approach to measuring GDP can be applied here, but it\’s a bit of a conceptual stretch. After all, advertising can be linked in a fairly direct way to the number of eyeballs or clicks, but the contribution that additional information makes in building up an overall database is harder to value:
\”The second avenue for the financing of free digital products is collecting and commercially exploiting the vast amounts of data generated by users of digital products. In many ways, this financing model resembles the advertising model: there is an implicit transaction between consumers (who provide data) and producers (who provide digital services for ‘free’ in return). A third party may or may not be involved. Economically speaking, the service provider finances its free services by building up a digital asset (volumes of data) that is subsequently used in the production of data services. … However (unlike the advertising model) the analogy is slightly more complicated here as there is no obvious proxy to establish the value of the services provided for free.\”
I won\’t try to do justice to their entire argument here, but a few other points are worth mentioning. There is a problem of valuing digital public goods, like Wikipedia or Linux. With conventional GDP, it\’s also difficult to value the 8 billion hours or so of volunteer time that Americans donate each year for other purposes. It seems clear that the value of \”knowledge-based assets\” is rising in companies, and for workers as well, and measuring the production and consumption of these assets is very hard. Digital transactions that cross international borders may cause ever-greater problems for GDP measurement, as well.
What seem to me the biggest challenges here are some classic issues for GDP statisticians that involve quality. Just to be clear, these issues of quality and variety aren\’t brand new in the digital economy. Even when just looking at goods, the many gradual improvements in quality can be very hard to capture. When thinking about services, the problem gets worse. When thinking about cost of a \”unit\” of health care services,\” or \”unit\” of banking service, or \”unit\” of legal services, it\’s quite hard to think about what the \”unit\” should be. In health care, for example, a day in an hospital, or a specific procedure like a colonoscopy, are quite different in their qualities now than they were a decade or two ago. Having dozens or hundreds of TV channels available is different in quality than than having only a few channels, just as the continual expansion of what is freely available on-line makes use of the internet a different quality experience.
The problems of measuring quality play out in a number of ways. When measuring output, an improvement in quality should should be viewed as a gain in actual real output, but it\’s not clear that the actual value of what is bought and sold captures that rise in value. An underlying problem here is that when it is hard to measure quality, it is also hard to measure prices and inflation. For example, the price of a day spent in a hospital room has risen dramatically over time. Presumably, some part of this increase is due to higher quality of what service is being provided, so it should be a rise in output. Indeed, perhaps the rise in the cost of a hospital room doesn\’t capture all of the rise in quality–so the rise in true output is actually bigger than the cost. Or perhaps some of the rise of the cost of the hotel room is just inflation. Economic researchers can make a career of delving into these kinds of issues, and the digital economy means that all the old questions need to be considered in new context.
However, there\’s one line which shouldn\’t be crossed. One sometimes hears the argument that the digital economy is understated in the GDP statistics because it doesn\’t measure the welfare or pleasure that people receive from various digital goods and services. But GDP is a measure of final goods and services bought and sold. GDP isn\’t welfare. It never has been welfare. To be sure, a high or a rising GDP is often correlated with many positive aspects of life for everyday people, But from the birth of the concept of GDP up to the present, no serious economists has ever argued that GDP is equal to welfare. Ahmad and Schreyer write:
\”[I]t is clear that consumer valuation should not attempt to measure total consumer welfare arising from the use of free digital products, just as the value of traditional market products is not a measure of consumer welfare. Measures of the total value of consumer welfare such as consumer surplus are at odds with the conceptual basis of measuring GDP and income, let alone any welfare measure that goes beyond consumption and encompasses quality-of-life dimensions. There is no question about the importance of such measures … However, measuring production and income is a different objective from measuring welfare.\”