People Lie, But Search Data Tell the Truth

Looking to Google for a revolution in social science.


Seth Stephens-Davidowitz, a former research assistant of mine, would not strike most people as a revolutionary. Yet in his new book “Everybody Lies,” he argues persuasively for a mutiny in social science.

The problem should be familiar to anyone who’s followed political polling in the past few years, despite the successful predictions of Emmanuel Macron’s victory in France. Put simply, most people tend to lie on surveys and on social media, too. As a result, when we study people’s responses to surveys or what they say on social media, we come up with a misleading picture.

Rather than disparage surveys and social media posts, Stephens-Davidowitz points to a different way of understanding ourselves. In the ostensible privacy of online searching, he argues, we inadvertently reveal ourselves, and this digital truth serum offers the best way of finding out who we really are.

Examples abound. According to survey data, Americans overall are not particularly racist, and any racism that does exist is more dominant in the South — a view that is often endorsed by the media. Yet online searches reveal a remarkable number of racist inquiries by Americans, and these searches are in no way limited to the South. Indeed, the highest rates for racist searches are found in places such as upstate New York, eastern Ohio and western Pennsylvania. The true racism divide is not North-South, it turns out, but East-West, with limited racist search behavior west of the Mississippi River. This pattern correlates strongly with presidential election results; in the local areas with the highest share of racist online searches, Barack Obama substantially under-performed, and Donald Trump substantially over-performed.

Another example involves homosexuality. Survey data and social media profiles suggest the proportion of men who report being gay is roughly twice as high in Rhode Island as it is in Mississippi. Yet Google searches of terms associated with gay pornography vary little across the country, and are only marginally higher in Rhode Island than in Mississippi — suggesting that the survey results and social media profiles in some states may not reflect reality. Indeed, in the states where under-reporting may be larger, spouses tend to be more suspicious. The most searched-for term on Google after “Is my husband…” is not “cheating” or “depressed” but “gay,” and that question is asked far more frequently in states where the survey reports are low.

Many other myths are exploded in the book, some by search data and some by other evidence. The notion that violent movies cause violence? Not correct. The crime data show that violence declines before, during and after the showing of violent movies — perhaps because people who would be inclined to commit violence instead go to see the violent movie, and given the association between drinking and violence, the diversionary effect lingers because movie theaters generally don’t serve alcohol.

Another accepted idea, first offered by the historian James McPherson, is that the Civil War caused Americans to shift common usage from “the United States are” to “the United States is.” Nope again: A search of digitized books shows that there was no noticeable shift around the time of the war, and “the United States are” remained common for 15 or more years afterward.

Consider, next, the assumption that people start out liberal and become more conservative as they age. Again, not really. Instead, what seems to matter is an imprint effect that occurs when people are 18. Americans born in 1941, for example, turned 18 during Dwight Eisenhower’s presidency. And by about 10 percentage points, they have tended to be lifelong Republicans. A similar phenomenon applies to sports teams: A person’s favorite baseball team tends to be one that won the World Series during his or her childhood.

All of this would be merely amusing if it left us with only a collection of punctured myths. But Stephens-Davidowitz aims higher, writing that “Google searches are the most important dataset ever collected on the human psyche.” Therein lies the power of his new book: While acknowledging the limitations, Stephens-Davidowitz argues that big data can rescue social science from its garbage in-garbage out problem.

We are still early on this journey, but “Everybody Lies” provides the ballast to suggest it’s the right road.

Originally published at on May 09, 2017.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.


A Better Goal for Trump on Health Care

Forget about repealing Obamacare. Work instead on lowering medical costs and improving treatment.


Rather than renew their failed effort to repeal and replace Obamacare, President Donald Trump and congressional Republicans should move on to another aspect of health care: the need to contain costs and improve value.

Such a shift would allow them to be far more productive. For many if not most Americans, cost trends and value matter more than what’s happening on the individual insurance exchanges. Progress on this front would raise people’s take-home pay and improve the nation’s long-term fiscal balance, while also constraining the growth in premiums for those who buy insurance on the exchanges.

All the recent debate over the individual insurance market has made it easy to lose sense of the broader picture. In 2015, more than 155 million Americans received health insurance through an employer, and another 43 million, through Medicare, according to the Kaiser Family Foundation. Roughly 80 million more received coverage through Medicaid or the individual market — the areas where Obamacare expanded access — but even here, most of the coverage still reflects pre-Obamacare Medicaid. Consider that repealing Obamacare would reduce insurance coverage by about 30 million people in 2026, according to the Congressional Budget Office. While that’s a very large number, the population with coverage through other sources is many times greater, and these Americans still spend too much for too little on health care.

So what could the Trump team do to improve value? The way to start is by addressing the extreme variation in health costs across the U.S. Within Medicare, most of the variation reflects the amount of care provided — especially in post-acute care (the care a person receives after he or she leaves a hospital, including in skilled nursing facilities). Within the world of employer-provided insurance, in contrast, most of the variation reflects prices paid.

The accumulated evidence suggests that more care is not better, at least on average. That has been confirmed by clever research just published in the Journal of Health Economics, which analyzed how costs and outcomes vary depending on which ambulance company happens to pick up a particular patient. It turns out that ambulance companies tend to funnel patients to particular hospitals, so the rotation among ambulance companies provides an almost random rotation of patients to different hospital systems. The results show that there is little if any connection between overall spending and subsequent mortality rates. But they also show that higher in-hospital spending seems to be associated with better outcomes.

In contrast, higher levels of spending after patients leave the hospital (the part that drives variation in Medicare costs) is associated with lower-quality care.

These results light the path forward for President Trump: Find ways to lower prices in employer-provided insurance, and to reduce utilization (especially of post-acute care) in Medicare. In future columns, I will explore each of these paths in more detail.

This agenda would also enable Trump to involve Democrats, whom he will need as part of a new governing coalition to pass spending bills and a debt limit increase this year. After all, plenty of Democrats would rather find ways to deliver better-value health care than debate proposals to take people’s insurance coverage away.

Originally published at on April 25, 2017.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

Management Practices Matter More Than You Think

To understand macroeconomic changes, look at how individual companies are run.

Management consultants face perennial questions about what value they add to companies. But management practices go a long way toward explaining why some businesses perform better than others, an important new analysisshows. Perhaps management consultants are onto something after all.

Surprisingly large and growing differences across businesses in wages, productivity, capital returns and worker mobility may influence income inequality and even macroeconomic growth, many recent studies show. Now it seems management practices play a big role in explaining the variations across businesses, at least in manufacturing.

The new study, by a group of well-respected researchers, is based on a Census Bureau survey of about 32,000 U.S. manufacturing plants. The survey asked such things as how frequently managers track performance indicators, how quickly underperforming employees are reassigned or dismissed, and whether managers are promoted based solely on performance and ability.

The researchers used the companies’ answers to construct a management practices index, with higher ratings for plants that do such things as monitor performance, detail targets and tie management incentives to performance. Because the survey included multiple plants within individual firms, the economists were able to examine how practices vary both within companies and between them.

They found, first, that management techniques vary widely from plant to plant. Less than 20 percent use three-quarters or more of the performance-oriented management techniques, for example, while more than a quarter use less than half of them. Perhaps most surprisingly, the authors found that a little more than 40 percent of the variation in overall management practices occurs within the same firms.

They also found that the management techniques matter — a lot. The plants practicing more structured performance-oriented management are more productive, innovative and profitable. Every 10 percent increase in a plant’s management index is associated with a 14 percent increase in labor productivity, for example. And the relationships hold over time: The more performance-oriented a plant becomes, the more productive it is. Companies with higher management scores are also more likely to expand and to survive.

The researchers were able to compare the management approaches with more traditional explanations of business performance — things such as research and development, information-technology expenditures and workers’ skill levels. The authors lined up plants according to total productivity, and looked at differences between those ranking in the 90th percentile and those in the 10th percentile. Management techniques can explain 18 percent of that difference, they found, while R&D accounts for 17 percent; employee skills, 11 percent; and IT variation, 8 percent. In other words, management matters more than conventional explanations for performance.

Finally, the researchers looked into why management practices vary so much. They examined factors such as the competitiveness of the market in which a plant operates, the business environment (including state Right to Work laws), whether there is a college nearby, and learning spillovers from large multinational plants. All these other factors matter, but collectively they explain only about a third of the variation in management techniques.

Whatever the larger explanation, management practices vary substantially, even within manufacturing companies, and they cause big differences in performance. Those differences, in turn, have macroeconomic implications. Someone worried about why wage inequality has risen in the U.S., or why productivity growth has declined, may not immediately think to question why some companies are well managed and others aren’t. But increasingly, the evidence shows that those questions matter.

Originally published at on April 11, 2017.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

No Raise? It’s Not You. It’s Your Company.

At top-performing firms, workers of all skill levels climb the income ladder.

The kind of company you work for makes a big difference to your chances of getting raises, new research has found. This adds to growing evidence that what goes on inside firms matters beyond their walls. Researchers have shown that company-level differences have become large enough to influence national productivity growth and overall wage inequality. The new study suggests they affect income mobility, too.

Having gathered data on workers and companies from the Census Bureau and the Social Security Administration, researchers John Abowd of the Census Bureau, Kevin McKinney of the California Census Research Data Center and Nellie Zhao of Cornell University categorized workers and their employers along three dimensions: skill, earnings and average company pay. Not surprisingly, they found some correlation between workers and firms. Low-skilled workers tend to work at low-paying companies, for example, and to earn low wages.

From Bloomberg View – Articles by Peter R. Orszag

The Wrong Way to Lower Health-Insurance Premiums

Republicans’ Obamacare replacement would bring lower-value policies.

For proponents of the American Health Care Act, perhaps the most encouraging nugget in the Congressional Budget Office’s otherwise critical analysis is that insurance premiums could fall by 10 percent on average by 2026. Even this prediction is more mirage than reality, however, in part because of an obscure concept known as “actuarial value.”

As many opponents of the Republicans’ Obamacare replacement legislation have already noted, for many people, the decline in premiums would be smaller than the cutback in their subsidies, so they would still end up paying more. And in any case, the predicted fall in premiums partly reflects a troubling rise in the share of older Americans without insurance, a change that would shift the enrollment pool to younger, less expensive beneficiaries.

From Bloomberg View – Articles by Peter R. Orszag

Selling Health Insurance Across State Lines Won’t Save Money

An idea that sounds easy is too complicated to work.

The effort to replace Obamacare faces increasing challenges, the more it is subjected to the harsh light of scrutiny. A good example is the proposal, apparently central to the Republican replacement plans, to allow people to buy health insurance across state lines.

This idea has been put forward as an elixir to all sorts of health sector problems. In his joint address to Congress, President Donald Trump argued that allowing people to buy health insurance in other states would “create a truly competitive national marketplace that will bring costs way down and provide far better care. So important.”

From Bloomberg View – Articles by Peter R. Orszag

Alexander Hamilton Loved Employee Ownership. So Should We.

By protecting jobs, worker equity plans support the whole economy.

Employee equity ownership of companies has been promoted in the U.S. since the country’s founding. After the Revolutionary War, Treasury Secretary Alexander Hamilton fostered the resurgence of the codfish industry by providing financial subsidies — but only for ships that had written profit-sharing agreements with their crews. This program, enacted in 1792, lasted until after the Civil War.

From Bloomberg View – Articles by Peter R. Orszag