Given the slow growth in wages over the last several decades, especially for the bottom half of the U.S. labor market, progressives are right to call for a higher minimum wage. The current federal level—$7.25 an hour—has not increased in eight years and is 35 percent lower in inflation-adjusted terms than it was in 1969, despite the fact that since then, U.S. productivity has more than doubled (up 135 percent).

The progressive case for a minimum wage—whether it’s an increase to $10.10 per hour, as President Obama once called for, or for an even higher increase to a “living wage” of around $15—is made almost exclusively on the premise of promoting greater fairness. Tens of millions of U.S. workers live in or near poverty and, if we could raise the minimum wage, their lives would improve in big ways. The Center for American Progress makes its case on the grounds that a higher minimum wage will help support “family values”; presumably they believe this argument will also help broaden their appeal among religious conservatives. In a blog post for the Center on Budget and Policy Priorities, Jared Bernstein states his support for a higher minimum wage because it will “help low wage workers make ends meet.” And writing for the Economic Policy Institute, David Cooper notes that the key reason to raise the minimum wage to $15 an hour is because that “directly lift[s] the wages of 22.5 million workers.”

While this line of argument no doubt rings true, when it comes to arguments about economic policy, growth, jobs, and efficiency most often trump fairness. In other words, while conservatives and economists of all stripes might acknowledge that a higher minimum wage will lead to fairer outcomes, many also warn that it will come at the cost of lost jobs and stunted growth. That almost inevitably beats fairness. Game, set, match—or at least, tie.

If progressives want to break through this frustrating stalemate and get a higher minimum wage over the finish line—at least in more states, if not in Congress—it’s time for them to make the case for a higher minimum wage on the grounds of growth first, and fairness second. In other words, not only should progressives stop ceding ground to opponents when it comes to jobs and GDP growth, they should rightly assert that a higher minimum wage would actually improve both. In other words: If we want to grow the U.S. economy, not just redistribute more of its fruits to low-income workers, we need to raise the minimum wage. This argument is much more likely to prevail.

Progressives need to move the needle in this debate beyond whether the minimum wage kills jobs in firms that employ minimum-wage workers. Interesting debate, but not to the point. The argument among most actual economists on this issue is fought out in the trenches (and in the weeds) of microeconomics. In other words, economists’ focus is on the impact of a higher minimum wage on individual firms and on the workers subject to it. For example, the classic study by David Card and Alan Krueger looked at the impact of an increase in the minimum wage on fast food restaurants in New Jersey and found that it had no negative impact on employment levels within these businesses. Likewise, although the study had some constraints, Arindrajit Dube, William Lester, and Michael Reich saw the same results across the entire nation. In other words, the firms they studied were either able to avoid raising prices due to the increased minimum wage, or if they did increase prices, those hikes did not reduce sales and employment. However, another study by David Neumark, Ian Salas, and William Wascher found that a higher minimum wage did result in fewer jobs in the affected firms.

So, not surprisingly, liberals point to Card and conservatives to Neumark. It comes down to dueling studies—our word against theirs. It’s hard to win this one; beyond these studies, the common sense explanation, for most people, is that a higher minimum wage would raise a firm’s costs, leading to higher prices, lower sales, and reduced employment. At best, the game is tied, and a tied game does not a minimum wage victory make.

The problem with these micro-level studies, though, is that society at large should be largely indifferent to effects on these individual firms. They might employ fewer workers, and they might not, just like they may or may employ fewer workers after installing robots, for instance. But what’s truly important is how many jobs there are in the U.S. economy after raising the minimum wage. And, on that point, the answer is clear: the exact same number as before, except that they pay better.

To accurately assess overall employment impacts, whether from robots or a higher price for labor, we need to look at second-order impacts. In the case of the minimum wage, an employer may or may not hire the same number of workers, but if they hire fewer staff because of higher labor costs, then it’s important to recognize that the remaining workers now earn more. These workers don’t bury their extra earnings under the mattress, nor are they likely to save much given that they have so little money to begin with. Rather, the workers spend that income on additional consumption: much-needed health care, repairing their car or buying one that actually works, or even purchasing a bit more food so their kids don’t go to bed hungry. This increased spending creates demand, which leads other firms to create jobs, offsetting any jobs lost in the firms now paying the higher minimum wage.

This leads to a related and key point: Job creation is determined in the realm of macroeconomics, not microeconomics. So when opponents of a reasonable increase in the minimum wage say something like “A fundamental law of economics—the law of demand—states that when the price of anything (including labor) increases, the quantity demanded will decrease,” they are wrong. In this case, they are thinking of products and services, not labor. If a company raises the price of a good or service, demand will likely fall and the firm will employ fewer workers. But economy-wide labor demand is fundamentally different; it is largely determined at the macroeconomic level and by what federal monetary and fiscal policy does.

To see why, let’s stipulate for argument’s sake that a higher minimum wage would lead to increased prices at Joe’s Pizzeria (since Joe has to pay his workers more), and that this reduces the demand for his pizza (more people choose to cook dinner at home). Therefore, Joe doesn’t need to employ as many workers. Those workers would likely be unemployed until they can find new jobs, just as the minimum wage opponents warned. And this would raise the national unemployment rate. But wait; could the argument truly stop here? What happens when the unemployment rate increases? Well, traditionally, the Federal Reserve then reduces interest rates, which spurs increased spending and investment.

This dynamic continues until the economy regains full employment. So even if there were a onetime economic “shock” from a higher minimum wage (a risk that could be reduced by phasing in an increase over three or four years, a plan virtually all progressives support), that shock would be addressed by easing federal monetary policy, and the economy would relatively quickly return to full employment. This, coupled with the fact that Joe’s remaining workers are now earning and spending more and that this spending spurs more job creation, explains why, in the past, when the minimum wage was higher, unemployment was often quite low. For example, when the minimum wage was about $11 per hour (in 2017 dollars) in 1969, the unemployment rate was just 3.5 percent—compared to 4.1 percent today. So, when opponents of a higher minimum wage argue that it will lead to job loss, what they really mean is job loss at particular firms, not fewer overall jobs in the U.S. economy.

This means the jobs argument is either wrong or beside the point. But what about growth? Ensuring that more low-income workers make a higher wage is indeed inherently fair, but it’s not growth, in and of itself. When progressives do discuss a growth effect, they usually channel John Maynard Keynes and make some vague and unsupported assertions connecting more consumer spending to GDP growth. But Keynes never made this argument. Only when an economy is in recession does additional, induced spending spur faster growth, by drawing underutilized economic resources into the economy. But in periods of full employment, more spending without an accompanying increase in the labor force or productivity just leads to inflation.

So, to effectively make the growth argument we must return to firms and workers. Let’s start with workers. While the current unemployment rate is low, the rate of labor force participation (the share of working-age adults working or looking for work) isn’t great: 63 percent, down from 67 percent in 1997. There are many theories as to why this is, but surely one reason is that, at such low wage levels, it is easier for a low-skill worker to choose to sit out of the labor market altogether. By making it more remunerative to work, a higher minimum wage will induce more workers, especially male ones, to return to it. Duke University researchers Tom Ahn and Peter Arcidiacono found that “An increase in the minimum wage induces some workers who were previously not searching to participate in the labor market.” Likewise, Princeton University economist Alan Krueger writes that “Policies that raise after-tax wages for low-wage workers, such as an increase in the minimum wage or expansion of the Earned Income Tax Credit, would also likely help raise labor force participation.” Even some conservative opponents of the minimum wage acknowledge this point. The Heritage Foundation’s James Sherk, for example, has stated that “Increased minimum wages encourage workers with higher reservation wages to enter the job market.” That increased participation, in turn, is what leads to an increase in GDP.

But there’s a second and even more important aspect of a higher minimum wage that will drive GDP growth: firm productivity. Let’s consider Joe’s pizza business again: It must now pay its delivery drivers and pizza cooks 25 percent higher wages. Because Joe has to pay more for labor, it becomes more economical for him to adopt technology, such as more modern ovens, to automate some of the work. The introduction of this kind of labor-saving machinery is a key driver of labor productivity, and the technology’s adoption by firms is driven in part by the cost of equipment relative to the cost of labor. When the price of labor is high, the return on investment from labor-saving technology is increased because such an investment saves the firm more. This exemplifies the “Webb effect”—the theory that a higher wage floor leads to higher levels of efficiency.

Studies confirm this effect. For example, MIT’s Daron Acemoglu finds that in the absence of minimum wage legislation, the U.S. labor market is inefficiently biased toward low-wage jobs. In a comprehensive analysis of the impact of the minimum wage on growth, economist Robert Prasch writes that “Firms faced with higher wages are forced to employ more advanced equipment.” Another study concludes that “If the federal government raises the minimum wage, employers in some sectors may expedite the adoption of automated equipment and new technology to increase labor productivity.” Therefore, a higher minimum wage makes it more economical for organizations to substitute capital for labor, typically low-wage labor. Even Trump’s original secretary of labor nominee, Andrew Puzder, agrees with this, having stated that, with a higher minimum wage, “Businesses reduce staff and automate.”

Automation is a key way the economy grows. Automation of the farm—tractors instead of horse-pulled plows—is why we pay so little for food. Automation of the factory—computer-based machine tools instead of files and saws—is why the price of goods has fallen in real terms over the last century. And automation of fast food restaurants and other industries that hire many minimum-wage workers is how we are going to continue to increase productivity and, hence, wages. That is another aspect progressives should cheer: The result of a higher minimum wage would not be just more jobs with higher wages, but also fewer jobs in low-wage occupations.

Unfortunately, U.S. businesses have been slow to adopt labor-saving technology, in part because of the low costs of unskilled labor in this country. This is one reason why the productivity growth of the last decade has been the slowest since the government began tracking it in 1947. Compared with other advanced nations, low-skilled labor is particularly cheap in the United States. The federal minimum wage of $7.25 per hour is far below many developed countries. For example, in 2016, the minimum wage in Australia was the equivalent of $11.10 per hour, $10.30 in Germany, $9.10 in Ireland, $9.90 in the Netherlands, and $8.40 in the United Kingdom (in 2015 prices adjusted for cost of living). Not surprisingly, these countries with higher wages are generally more likely to adopt labor-saving technology, like different self-service options. For example, you are unlikely to see a parking garage attendant in the Netherlands, as garages mostly use self-serve payments. Likewise, chip payment cards are mandatory on all public transit, and kiosk ordering in fast food restaurants is common.

Of course, at this point, opponents of a higher minimum wage say “We told you so”; labor-saving technology will lead to fewer jobs, so the minimum wage is a job killer. But this is even more fallacious than the notion that a higher minimum wage itself reduces employment. Why? Because, when firms reduce costs through automation, those savings raise wages or lower prices, or both. In any of those three scenarios, those savings go back into the economy, creating other, often better jobs. So, if opponents of the minimum wage want to claim their opposition on the basis that it will lead to automation, then they need to get out there on the barricades and fight the robots, too. Maybe they can even join in with those calling for taxing robots?

Given the available evidence, why do so many economists continue to oppose the minimum wage? They do so because conventional, “neoclassical” economists live in a world where virtually any government intervention in the economy is seen as distorting the “free market.” They base their opposition on the idea that if an employer and a worker voluntarily agree on a wage, then that wage level must be welfare maximizing for both parties, and by definition, for society at large. If government interposes itself between these two willing parties with a mandated minimum wage, all it can do is distort labor markets and lead to less, not more, economic welfare. American Enterprise Institute scholar Mark Perry sums up the view neatly here: “Minimum wage laws prevent mutually advantageous, voluntary labor agreements between employers and employees from taking place.”

But this tidy textbook worldview is grounded on several heroic assumptions that simply are not true in the real world. One is that firms are rational when it comes to adopting new capital equipment. In fact, similarly situated firms differ significantly in their willingness to buy new capital equipment. A second assumption is that there are no external benefits (what economists call externalities) from a firm investing in capital equipment. In other words, many economists believe that whether firms adopt more and better capital equipment is of no concern to public policy because the benefits only accrue to the firm itself. There are no externalities. In fact, a slew of recent studies shows that, on average, firms don’t capture anywhere close to all the benefits of their investments in new machinery; these benefits inevitably spill over to competitors and to society at large. But in the absence of incentives for firms to invest more, such as a higher minimum wage, firms will underinvest in new equipment relative to what is societally optimal.

This relates to another key failure of the free market: the fact that it’s possible for economies to be in perfect equilibrium (the nirvana state for economists) in two fundamentally different ways, as relates to skills and investment. Research by economist Elvio Accinelli and colleagues has shown that there is strategic complementarity between the percentage of high-skill workers and high-value-added, innovative firms in an economy. Their research finds that economies can be in perfect neoclassical equilibrium at either a high level of innovation, high skill levels, and higher wages, or in a “poverty trap” of low skills, low wages, and underinvestment in new capital equipment. In other words, if there are not enough skilled workers, firms will not adopt advanced technology that leads to higher productivity since their workers don’t have the skills needed to operate it. And if firms don’t adopt advanced technologies, workers won’t seek out the skills needed to use them. The “poverty trap” can therefore be avoided if firms have stronger incentives to invest in new capital equipment, while at the same time society increases its efforts to boost worker skills.

Unfortunately, from the perspective of the individual firm, the low-wage, low-investment path can be quite profitable. As Susan Helper and Ryan Noonan found in a study for the U.S. Department of Commerce, firms can use either one of these “production recipes” to be profitable. Helper and Ryan find that these practices are highly correlated; firms adopting one are more likely to adopt the others. For example, there is a positive 0.7 correlation between payroll per employee and value-added per employee (in other words, firms that pay their workers more have higher productivity), and a 0.6 correlation between payroll and capital expenditures (firms that pay their workers more invest more in capital goods).

It should come as little surprise that the U.S. economy has fallen into this second order of things, the “poverty trap.” There are too many low-wage, low-skill jobs, too little investment by companies in new machinery and high-performance work organizations, and too little support by government for those organizations, including skills development. Getting out of this trap will require a wide range of policies, including better programs to boost worker skills. But no policy change is more vital here than a higher minimum wage. And, as such, progressives will need to champion such a move, by highlighting the essential role it will play in creating a robust economy and growth for all.