Why aren’t new technologies making us more productive?

For many years, large American companies firmly believed that cloud computing and artificial intelligence would fuel a surge in productivity and thus wealth creation. This conviction justified a deluge of investments by the sector of capital for ventures, and of expenditures by companies. And the results, the proponents insist, will not be confined to a small group of tech giants, but will be felt throughout the economy.

None of this has happened so far.

Productivity, defined as the value of goods and services produced per hour of work, dropped sharply in the first quarter of this year, the US government announced this month. The quarterly numbers are often volatile, but the report appeared to shatter earlier hopes that a pick-up in productivity growth had finally arrived, helped by accelerated investment in digital technologies during the pandemic.

Productivity growth since the pandemic erupted is now at about 1% a year, which is in line with the mediocre average prevailing since 2010 — and well below the last period of robust improvement in productivity, between 1996 and 2004, when its growth was more than 3% per year.

Economies grow not just by adding more capital and more labor. Another vital ingredient is a country’s ability to create and commercialize innovation, which makes investment and workers more productive.

Seemingly small percentage gains in productivity can make a big difference to a country’s wealth and standard of living over time. An additional 1% annual increase in productivity for a few years — say until 2024 — would generate an additional $3,500 per capita income for Americans, the McKinsey consultancy estimated in a report last year.

The average annual increase of 3.8% in productivity between 1948 and 1972 was the engine of post-war prosperity in the United States.

Inequality perpetuates itself with low productivity

Productivity is not a cure-all for economic ills. “Even if optimism about the current wave of digital technology proves justified, that doesn’t mean the benefits would actually be shared,” said Laura Tyson, a professor at the Haas School of Business Administration at the University of California at Berkeley and chair of the White House Council of Economic Advisers under Bill Clinton’s presidency.

But a less productive economy is a smaller economy, and one with fewer resources to deal with social challenges such as inequality.

The current productivity conundrum is a subject of vigorous debate among economists. Robert Gordon, an economist at Northwestern University, is one of the most skeptical analysts of the new technology.

Today’s artificial intelligence, he said, is primarily pattern recognition technology, and it sifts through vast repositories of words, images and numbers. His exploits, according to Gordon, are “impressive but not transformative”, unlike, for example, electrical power and the internal combustion engine.

Erik Brynjolfsson, director of Stanford University’s Digital Economics Laboratory, is a leader in the field of optimists. He confesses to feeling a little disappointed that the increase in productivity has not yet been confirmed, but he is convinced that it is only a matter of time.

“There are real changes happening. A tidal wave of transformation is underway,” said Brynjolfsson. “We’re seeing more and more hard evidence.”

It will likely take years for definitive answers to emerge in the productivity debate. Brynjolfsson and Gordon made a “long bet” last year, the winner of which will be decided in 2029. But industry- and individual-company-level studies, leveraging data that derive from business census results and online job postings, demonstrate the pattern. diffusion of technologies, and the obstacles they face.

The leaders are mostly companies that have been investing in digital technology for years and younger, high-growth companies, many of which are incorporated with support from the venture capital sector. Cloud computing has been widely adopted, but more advanced technologies such as artificial intelligence applications are not yet.

The limited adoption, some experts say, isn’t all that surprising at this point, given that three-quarters of US businesses are small, with fewer than 10 workers.

Revolution in teleservice

At Anthem, a health insurance company whose plans cover more than 45 million customers, around 75% of consumer questions are now answered through digital channels, including a web portal, a mobile app and voice recognition software. Three years ago, the share of digital service was around 30%.

The question-answering technology to help people with basic tasks like checking the status of a claim, paying a monthly fee or finding a doctor is partly powered by artificial intelligence.

Digital automation eliminated 10 million phone calls that Anthem’s call centers would have had to answer, estimated Rajeev Ronanki, president of digital platforms at Anthem.

Anthem, which next month will change its name to Elevance Health, has not reduced its customer service team. But the role of the workers who are part of it and the way in which their performance is evaluated has changed. The traditional indicator of call center performance is “call handling time”, and the less time the better.

Anthem now wants its teams to resolve customer issues with one phone call whenever possible, rather than passing them on to another department.

Many of the agents in the company’s call centers received additional training to become what the company defines as “service navigators.” Indicators of their performance now include resolved issues and customer satisfaction surveys.

Under this broader set of metrics, Ronanki said, the company’s agents became between 30% and 40% more productive. Adding capabilities and redesigning the work, he said, is just as important as improving technology.

“Building technical capability is just the beginning,” said Ronanki.

Translation by Paulo Migliacci