For better or worse, the pandemic changed the economy in many ways. To move forward, we need to address the issues raised by the altered business and consumer patterns. But before public and private sector policies can be designed, it is important to actually know who did well and who needs support.

When the pandemic hit, economists puzzled over the potential shape of the recovery. Since, as Bernard Anderson, former United States assistant secretary of labor, commented, “The pandemic created conditions that upset normal, cyclical labor market adjustments.”

There was the hockey stick or swoosh model, which assumed slow but steady improvement. The “V” proponents argued that growth would rebound rapidly, while the “U” backers thought there would be a period of slow growth followed by a sharp upturn. Some even thought we could see a “W”: a tentative recovery followed by a second mild downturn and then the real recovery.

But there was one model that broke the mold. It was called the “K” recovery, which postulated that some sectors of the economy and the workforce would recover steadily if not solidly, while others would decline for an extended period.

» READ MORE: Rental car prices skyrocket as the pandemic causes supply shortages

Which best describes the recovery? It depends on the data you use for the analysis.

If GDP is the measure, the “V” recovery looks like the winner. Second-quarter 2021 economic activity exceeded the previous high in the fourth quarter of 2019.

But if you use employment, the picture changes.

In August, there were still more than five million fewer employees on the nation’s payrolls than in February 2020. The number of unemployed remains 2.7 million above its low point and the unemployment rate is 1.7 percentage points higher than its early 2020 bottom. The labor force participation rate is down sharply, as five million people have dropped out. So, maybe either the hockey-stick or swoosh best describes the recovery.

But these approaches use the broader macro-level data. The problem is, it’s the more granular details that matter as they tell a more complete story about who are the winners and losers.

The demographic data show the need to go deeply into the numbers. Since February 2020, the U.S. labor force has shrunk by about three million people. Women 20 and older comprised 55% of that total decline, despite the fact that they make up only about 45% of the workforce.

For Black women, it’s more distressing. They make up about 8.5% of the total workforce but accounted for more than 12% of the decline. Some have done well, but others haven’t. The “K” recovery may be the most accurate.

Differences across sex and race, which are just one example, are critical for policy. Before funding programs to ease the economic impact created or exacerbated by the pandemic, you must first identify the problems, one of which is the uneven impact on the labor supply of women and especially women of color.

It is critical that labor shortages are limiting the ability of businesses to meet growing demand. The single most important issue is labor, according to the National Federation of Independent Business’s July Small Business Economic Trends report. Small businesses, a major engine of job growth, face an inability to hire qualified labor.

A major priority should be expanding the labor force.

And where is the problem greatest? Women, and especially minority women, who have seemingly been forced out of supplying their labor. That is why many are pushing for fundamental changes in child care, which tends to fall most heavily on women.

Similar disparities exist across industries and even within the same industry. Some, such as those related to internet activities, remote businesses, technology, goods transportation, certain manufacturing companies, and construction saw their businesses rebound sharply. But others, including hospitality, continue to fight the cautiousness created by the lingering pandemic, as well as the changes in spending patterns and higher operating costs.

But the greatest issue may be in determining which workers have been hurt the most. Here, the data are clear: As a recent report by the U.S. Bureau of Labor Statistics points out, “establishments paying the lowest average wages and the lowest wage workers had the steepest decline in employment and experienced the most persistent losses as of the last months of our analysis…” (https://www.bls.gov/osmr/research-papers/2021/pdf/ec210020.pdf)

Unfortunately, not all the data we need to definitively determine the losers vs the winners is available. It may be months before we know which employees and firms need the greatest help. Even saying that low-income workers should be the target is not enough, as many of the industries, such as hospitality, that employ those workers have the highest rate of job openings.

We need to have some idea about what it will take to get workers to accept or return to jobs in retail or restaurants if we are going to help both the businesses and the workers, but right now that is not clear. Again, quoting Anderson: “There was a structural shift in the labor/output ratio that will take time to reverse, and probably will not return entirely to the pre-pandemic conditions.”

As the government puts together its next set of economic policies, it is not enough to argue that what worked in the past (e.g., traditional infrastructure projects) should take precedence over what is being called “soft-infrastructure,” such as child care. Conditions have changed and policies need to be updated so they target the current problems.

Joel L. Naroff is the president and founder of Naroff Economics, a strategic economic consulting firm in Bucks County.