Can we see the computer age in the productivity statistics?

Question

Robert Solow famously said:

You can see the computer age everywhere but in the productivity statistics

This quote is often linked with the productivity paradox (https://en.wikipedia.org/wiki/Productivity_paradox), such that we don't see mayor productivity growth from new tech.

But there is something I don't understand. Economic growth is about percentage growth. A graph of GDP looks like an exponential function. So nominally a 5% growth is much much higher when the economy is that of the US than that of Malawi.

So we might not see an acceleration of economic growth, but we at least see continued growth, which allegedly becomes harder and harder as economies develop. This is the idea behind Solow's model, where without an exogenous percentage growth in A, output converges to the steady state.

So I don't really understand the problem. To me, computers are clearly visible in a growing economy. But this seems to be unsatisfactory for Solow, who wants to see what, a 10% growth rate?

What's wrong in my thinking?

Answer 1

It means that computers (or IT in general) contribute to the society (they make us better off and more productive at work), but this does not show up in GDP statistics. The reason is that many of these services are for free. As an example, the value of the Google Search is not reflected in a market price for running this search. The problem is related to the societal value of mothers raising their children at home. Other examples are environmental protection or caring for your parents. Since these services are not traded on markets, it does not enter GDP. Whether the Solow quote is actually true is however debatable. The value of the google search engine is to some extent reflected in the advertisement google is selling. And this enters GDP statistics. The relation with the productivity puzzle is whether productivity is actually falling or whether we need new concepts to measure productivity in the computer age.

Answer 2

Solow's Model - and much of growth theory - is somewhat based on resource saving technical change: economies get more productive by producing more output with the same amount of inputs.

This is a very useful framework in a "steel and wheat" economy, but loses explanatory power when you want to account Google Translator and similar services which are for free, available for anyone at any time.

One response is to think about output-saving technical change, as in Hulten and Nakamura, 2017:

We extend the conventional neoclassical production and growth framework, with its emphasis on total factor productivity as the primary macroeconomic mechanism of innovation, to allow for technical change that affects consumer welfare directly. Our model is based on Lancaster’s “New Approach to Consumer Theory,” in which there is a separate consumption technology that transforms goods, measured at production cost, into utility. This technology can shift over time, allowing consumers to make more efficient use of each dollar of income. This is an output-saving technical change, in contrast to the resource-saving technical change of the TFP residual. The output-saving formulation is a natural way to think about the free information goods available over the Internet, which bypass GDP and go directly to the consumer. It also leads to the concept of expanded GDP (EGDP), the sum of conventional supply-side GDP and a willingness-to-pay metric of the value of output-saving innovation to consumers. This alternative concept of GDP is linked to output-saving technical change and incorporates the value of those technology goods that have eluded the traditional concept. It thus provides a potentially more accurate representation of the economic progress occurring during the digital revolution. One implication of our model is that living standards, as measured by EGDP, can rise at a faster rate than real GDP growth, which may shed light on the question of how the latter can decline in an era of rapid innovation.