Desktop version

Home arrow Communication

  • Increase font
  • Decrease font

<<   CONTENTS   >>

Productivity and Income Inequality

Two important questions, for workers, arise from the digital revolution. One, how is digitization impacting income inequality? If the talents of individual workers become instantly available on a global basis, will there be a super-star phenomenon where income is concentrated in a few talented individuals? On the other hand, widely disbursed information about jobs in the long tail is opening up opportunities and hence decreasing income inequality. The super-star effect may dominate where the skill set is one-dimensional, like a star basketball player or software engineer. The long-tail equalizing effect may be more important in jobs where “soft” (not easily quantified or measured) skills along multiple dimensions are required. The super-star phenomenon has far-reaching policy implications and circles back to the employment statistics cited at the beginning of this chapter. Shortage of super-star skills relative to the soft skills is what is giving rise to the 1 % phenomenon. Aghion et al. write “there is a whole empirical literature that has flourished over the past twenty-five years which shows that productivity growth is a main component of total growth, that innovation is a key driver of productivity growth, particularly in advanced economies” [50]. They go on to note that “if we look at innovation (measured by the annual flow of patents) and top income inequality (measured by the top 1 % income share) in the US since 1960, we see that these two economic variables follow parallel evolutions.” In other words, innovation and inequality move together. Furthermore, innovation also gives rise to increased social mobility as “more innovation implies more creative destruction, i.e. more scope for having new innovators ... replace current firm owners.” The example of California is cited - it has the highest 1 % income shares and the highest levels of social mobility among all states in the US [50].

The second question concerns worker efficiency. Why hasn’t the digital revolution shown up in productivity statistics? The annual percent change in non-farm business sector productivity has been less than 1 % since 2012 (1.09 % in the first quarter of 2012, -0.13 % in 2013, 0.4 % in 2014, 0.71 % in 2015), decreasing to 0.67 % in the first quarter of 2016 [49]. There are three possible reasons.

First, perhaps this is a measurement problem since output is not consistently represented. For example, how does one account for free blogs that provide valuable information to users (content) but are written by unpaid authors? A lot of output now centers on customer experience, as in Apple’s ecosystem, which is intangible and doesn’t show up in output numbers. A more fundamental measurement issue arises due to the changing organizational structures of firms. This is reflected in the trend toward the “on- demand” economy where P2P markets are operating. Granularity in firms and products leads to many important data points being lost since new organizational forms are slipping under the old radar screen.

Productivity mismeasurement could account for some of this slowdown. Byrne et al. [51] address this question and while they “find considerable evidence of mismeasurement, [they] find no evidence that the biases have gotten worse since the early 2000s.” They account for intangible investments; quality-adjusted prices for computer hardware, communications equipment, specialized information-processing equipment and software; globalization; technical innovations in oil and gas production; and a new methodology to account for free digital services such as search and social media. They make the case that consumption of services such as social media, that substitute for other leisure activities, are “properly thought of as a nonmarket production.” While these services add value to households, their additional value from advertising-related revenue is small.

Byrne et al. conclude,

thus, there is probably some (at this point very, very small, but likely growing) downward bias in the growth rate of real GDP from the emergence of the sharing economy. It would be useful to have official statistics on the nominal output of the various types of services included in the sharing economy. [51]

It is harder to evaluate changes in overall welfare. There have been value increasing gains to households due to the Internet, but most of these gains are outside the purview of market-sector gross domestic product (GDP) and proposals to include them are debatable. Nevertheless, according to Byrne et al.

[A]vailable estimates of the welfare gains (based on the value of leisure time)

suggests that “free” digital services add the equivalent of perhaps 0.3 % of

GDP per year to wellbeing. That is small relative to the 1.75 % slowdown in

labor productivity growth in the business sector from 2004-2014.

Second, productivity enhancing innovations are diffusing slowly through the economy, falsely suggesting a decline in economic dynamism - both in the form of fewer startups and slower reallocation of labor resources in response to productivity differentials across sectors [36]. Innovation diffuses more slowly across the broad economy due to the requirement for available complementary technologies. Implementation of a new innovation may take time and hence not show up in productivity statistics. For example, despite the earlier invention of the printing press, a good postal system was crucial for spread of printed matter across most of Europe. Not until the Tassis brothers organized the courier system much later in the fifteenth century, linking important cities in the Holy Roman Empire, did the impact of the printed technology materialize [39]. Today, users may need to figure out the various productivity enhancing aspects of their devices, like the iPad, beyond the stated function. The functionality of new products and services is often developed by users, who customize the product’s functions to their own requirements. While the iPad was introduced in 2010, its present day use in restaurants and pre-schools is an unforeseen and, perhaps, fortuitous development.

This slow diffusion of technology shows up in the weak productivity statistics. Structural changes in the economy precipitated by shifts in the labor market from more productive to less productive sectors can give rise to a slower growth rate of productivity despite technological advances in many industries. Dani Rodrik says “This perverse outcome becomes possible when there is severe technological dualism in the economy and the more productive activities do not expand rapidly enough.” This “growth reducing structural change has been happening recently in the United States” [52]. For example, the Food and Beverage industry’s share of full?time equivalent workers grew by 0.22 % over 2010-2014 but declined in productivity over the same time period, while Broadcasting and Telecommunications (B&T) saw its employment share shrink by 0.16 % but its productivity grew [53]. In fact, productivity growth weighted by the industry’s share of full-time equivalent workers is highest for B&T among all industries. Disconnecting ownership from consumption in the B&T industry, an issue I discuss in Chap. 5, may have enabled more rapid diffusion of DCT in the broader communication, information, media and entertainment (CIME) industries.

<<   CONTENTS   >>

Related topics