By Erik Sass
In an age when technology carries more than just a whiff of Orwellian associations, “big data” was probably not the best nickname to give to the masses of information about us and everything else now collected by the Internet and digital devices and services of all stripes; maybe “super data” or “awesome data” would have been better. But the choice has been made, and anyway the real question is: does big data even matter? Or is it more a case of big data, big whoop?
No question, some extravagant promises have been made on its behalf, in areas ranging from marketing and customer service to industrial production and security threat analysis. For example, business consultancy McKinsey estimates that retailers using big data could increase their profit margins by 60%, while the U.S. healthcare industry could create $300 billion of added value every year if it were to make maximal use of big data, reducing total American expenditure on healthcare by 8%.
Thus it was interesting, and a little jarring, to hear a skeptical stance on big data presented at the Free Market Road Show’s conference on the topic at the University of Economics in Prague on April 4, 2019. The contrarian view came from Lukas Kovanda, chief economist with the Czech Fund and a lecturer at the university, who argued that after a decade of hype there just isn’t any, well, data to support the contention that Big Data is transforming the economy – yet.
Kovanda focused in particular on the contention that Big Data boosts productivity, allowing labor to do more, faster. The fact is productivity growth has fallen steadily in recent years, reaching levels not seen for over a century, Kovanda noted: “During the last 10 to 15 years, we have seen the lowest growth of productivity since the second half of the 19th century. For all of the 20th century the productivity growth was much higher… For me it’s a worrying sign. We have all this new stuff, new visions, new technologies, but still [economies] have been showing a very modest growth rate.”
In fact the phenomenon is so widely acknowledged that it has its own nickname among economists, the “productivity puzzle.” To cite just a few examples, British productivity rose just 0.5% in 2018, its worst performance in five years, while after a decade of weak performance overall Eurozone productivity growth actually went into reverse with a 0.4% contraction in the fourth quarter of last year. In the U.S. growth was a bit faster but still relatively modest, with a 1.3% increase in 2018, up slightly from 1.1% in 2017.
Searching for explanations, Kovanda presented two hypotheses. The first, more optimistic interpretation is that “We are now in kind of an installation phase. We try to adapt ourselves to new technologies, and we’re not able to adapt [right away], but maybe we can do it later.” The second, less cheery view is “the new technologies are not so productive like the [new] technologies of the 20th century, like combustion engines, airplanes, automobiles.”
There is at least some reason to hope that the first explanation will prove correct in the long run, Kovanda went on, recalling that with regard to “the inventions of the second half of 19th century… the full impact of these technologies we saw in the middle of the 20th century, after 50 years or so.” The implication, he noted, is that transforming big data into productivity gains may be “a task for many generations, or two or three generations ahead of us.”
At least one of Kovanda’s fellow panelists agreed that, even after a decade, big data is still really just getting started. Milan Slapak, president and CEO of GE in the Czech Republic and Slovakia, was bullish on big data’s long-term potential and stated that he was very pleased with GE’s productivity performance. However he agreed that overall “industrial productivity growth is at a record low” and also conceded that “these new technologies are maturing, they are not in their adult phase.” Slapak made it clear the new technologies are a longer-term play, “But if you don’t do it now, those productivity gains will be [even more delayed].” Slapak also argued “it’s a generational thing… The learning curve, and the implementation curve will get steeper, once a certain generation hit a senior leadership positions.”