Every fiscal quarter, we look to the federal Bureau of Economic Analysis for the latest data on the total income and output of the United States. The country’s headline indicator of economic conditions is Gross Domestic Product (GDP) — the sum total of all that’s produced in our economy. This number drives the national conversation about economic growth, but it provides precious little information about the financial security of most Americans.

If you pay attention to headline fundamentals only, it would certainly seem like we’re now in the best of times: growth hummed along at a pace of 2% in the second quarter of 2019, payroll employment has now grown for 105 straight months and the current expansion is the longest in U.S. history.

However, relatively buoyant headline indicators may belie the fact that many Americans face economic hardship. According to the Federal Reserve, 40% of U.S. households cannot afford a $400 emergency expense (yes, you read that correctly). In addition, more than 25% of U.S. households have no retirement savings whatsoever.

Why the mismatch? It comes down to the fact that how we measure economic activity matters — for what we measure is often what we find. If we use the wrong measurements, our metrics cannot give us meaningful clues about the economy, and neither can they buttress the best data-driven policy.

It turns out that our current growth analysis practices hearken back to the early 1930s, when the economist Simon Kuznets developed the first estimates of aggregate national income at the request of the Department of Commerce. The data from that report provided Congress and the president with the first good snapshot of economic activity and led to the introduction of national income and product accounts (NIPA) statistics in 1942. Nearly 30 years later, Kuznets received the Nobel Prize in economics “for his empirically founded interpretation of economic growth.”

At the time, Kuznets’ metric reflected economic progress throughout the country fairly comprehensively. Between 1963 and 1979, average national income growth was relatively close to the income growth observed by most typical Americans. Over that period, national growth averaged about 1.7%, and most Americans’ incomes grew at approximately the same pace. In fact, incomes in the bottom half of the distribution increased at a much quicker rate (2.6%) than those at the tippy top (about 1%).

However, since then, the paradigm has shifted. According to the Washington Center for Equitable Growth, from 1980 to 2014, pre-tax income growth for the top 1% of all earners was 204% in the United States, far above the national average of 61%. Over the same time period, pre-tax income for the bottom 50% of earners grew by just 1%.

Therefore, the official GDP growth statistics may be out of step with how most of us are really doing. Because they report average income across everyone in the United States, as incomes at the top skyrocket, they drag the average up with them, making the official growth measure less and less representative. As Paul Krugman, a Nobel Prize-winning economist, once wrote in a New York Times column, “If Jeff Bezos walks into a bar, the average wealth of the bar’s patrons suddenly shoots up to several billion dollars … But none of the non-Bezos drinkers have gotten any richer.”

Kuznets himself knew that his metric was imperfect — in his original report to Congress, he wrote, “The welfare of a nation can, therefore, scarcely be inferred from a measurement of national income.” So, what can be done?

We need updated, more comprehensive metrics to fit the changing times. A good start would be to disaggregate national income accounts data to show how those at all income levels (perhaps delineated by deciles) are doing quarter-to-quarter and over longer time horizons. The economists Thomas Piketty, Emmanuel Saez and Gabriel Zucman have done much of the leg work to build this metric already, finding that between 1980 and 2016, national growth averaged approximately 1.3%, yet the top 1 percent saw its incomes rise 2.9% and the bottom 90% only posted 1% growth.

Importantly, we need the Congress to require the Bureau of Economic Analysis to require quarterly reporting of disaggregated income growth. This will focus the attention of policymakers and financial markets on the need for widely shared growth, and fixing the way we measure economic activity will allow us to demand better fiscal policies — and to hold lawmakers accountable for said policies.

Moreover, in an election year with more progressives in the race than one can shake a stick at, we often hear candidates promise to act as a “champion for working people” and enact policies that will sustain economic growth. That’s all very well and good, but how will we know whether those policies are actually helping the right people if we cannot adequately measure their impacts? Even the best policies are only as good as the data and metrics underlying them.