The most disturbing economic trend today is the falling share of national income—the total amount of money earned within the country—going to workers. According to the Bureau of Economic Analysis (BEA), only 61.8 percent of national income went to compensation of employees in 2012, compared with 65.1 percent in 2001. (Historically, about two-thirds of national income has gone to employee compensation, which includes wages and salaries as well as supplements such as pension contributions and health insurance.) Since the vast majority of workers are in the middle class, this means the middle class has been falling behind over the past decade at an alarming pace.

The flip side to this trend is the rising share of national income going to capital—interest, rent, dividends, and other forms of so-called unearned income. Corporate profits have risen to 14.1 percent of national income from 8.5 percent in 2001. (Historically, corporate profits have been about 9 percent of national income.)

For some time, this trend was thought to be temporary—as with all economic data, these numbers fluctuate from year to year based on the business cycle, changes in inflation, interest rates, and other factors. It takes time for economists to conclude that a structural shift has occurred that puts a historical trend onto a new trajectory.

It now appears undeniable that just such a structural shift has indeed occurred. In a recent report from the Federal Reserve Bank of Cleveland, economists Margaret Jacobson and Filippo Occhino examined not just the BEA figures cited earlier but other data from the Bureau of Labor Statistics and Census Bureau to see whether they confirm the aggregate trend. They do. Jacobson and Occhino project that labor’s share of national income will stay up to 1.5 percentage points below its predicted long-run trend for the foreseeable future. That means $200 billion less income going to workers this year alone.

There are two major concerns about the downward trend of employee compensation. First is equity. The distribution of capital income is far more unequal than the distribution of wage income. The shift from labor income to capital income, therefore, will necessarily worsen income inequality, which is already severe.

The second problem is macroeconomic. Consumption represents about 70 percent of GDP; at least in the short run, economic growth is almost entirely a function of spending by consumers. And, obviously, spending is a function of income.

When faced with a shock to their wages resulting from a recession and higher unemployment, consumers can maintain their consumption to some degree by drawing down savings, going into debt, and receiving supplements from government in the form of unemployment compensation. But these things are of finite duration. At some point, consumption must be cut to a level consistent with consumers’ “permanent income,” as Milton Friedman called it.

As yet, there is no evidence in the aggregate data that consumers have started adjusting to a permanently lower level of aggregate income. When they do, it must necessarily reduce GDP growth unless consumer spending can be replaced by some other form of spending. The only three options are increased private-sector investment spending, a rise in net exports, or growth in government consumption or investment spending.

Given the current low level of interest rates, private investment spending could pick up a bit, but this may only exacerbate the decline in labor’s share of income. That is because much business investment today is going into labor-saving technology or automation. That yields higher labor productivity, but few jobs. Those workers that succeed will be paid more but will require more education and training to obtain jobs. And there won’t be enough such jobs to go around. Net exports may also rise, but their growth will be hampered by continuing economic problems in Europe—historically, the major market for U.S. exports.

Theoretically, government could step in to the breach and supplement labor’s share of income through redistribution, direct job creation, improved education, infrastructure spending, and so on. But these are exactly the same things that economists advocate whenever there is a severe economic downturn—often derided as “Keynesian economics.” We have just come through the second most severe recession in American history and yet there was deep resistance to any countercyclical spending whatsoever—back in 2009, some of Barack Obama’s economic advisers advocated a countercyclical program more than twice as big as the $787 billion package that was enacted, and even that just barely passed Congress.

The point is that if Congress was reluctant to enact a temporary government program to aid workers when the need was obvious and overwhelming, it is impossible to conceive that it will be willing to enact permanent measures to supplement wages or permanently raise the number of people working for government. Even some “liberals” who might be sympathetic to such measures are as obsessed with austerity and deficit reduction as conservatives, differing only on the mix of taxes and spending cuts needed to achieve a “sustainable” deficit.

It nearly goes without saying that the budget deficit, insofar as it arises from government spending on goods and services, supplements aggregate spending and prevents the decline in labor’s share of income from being even worse than it already is. But we have seen in these last few years enormous pressure to lay off government workers and cut their wages and benefits, and it’s hard to imagine that this won’t continue well into the future.

There is at least the possibility that tax reform could ameliorate labor’s condition. In a recent article, legal scholars Mary Louise Fellows and Lily Kahng laid out a strategy for accomplishing this.

They start from the idea that spending by businesses is favored by the tax code whereas spending by workers is not. The tax code assumes that all spending by a business is in pursuit of profit and therefore tax deductible. Spending by workers, however, is assumed to be ordinary consumption and not deductible except to a very limited extent, such as when a worker pays out of her own pocket for education to improve her skills in her current field. In contrast, education or training outlays to go into a new field are not deductible.

In the past, when almost all workers made their living working for wages paid by a corporation, the existing tax treatment was defensible. But today, when the traditional corporation is a declining form of business organization, and when workers are increasingly self-employed contractors, Fellows and Kahng argue that the tax code undertaxes capital and overtaxes labor.

They urge tax reforms that would curb business deductions while expanding those for workers that are necessary for them to be able to work. One place to start would be the lavish benefits paid to corporate executives that really just allow them to consume on the corporation’s dime, such as write-offs for corporate jets that executives use personally. Workers should have more latitude to deduct nonreimbursable business expenses, such as child care and commuting costs.

But such suggestions are just the beginning. We need a much deeper national conversation on the declining share of income going to labor and what to do about it. Tax policy is only part of the answer. International trade, labor unions, government employment, and many other areas of public policy must also be examined.

As long as capital can easily move across national borders to seek the highest return, and multinational corporations are largely beyond the reach of national governments, workers in high-wage economies will suffer. While some liberals believe that revitalizing labor unions can ameliorate these problems, that is a pipe dream; unions are declining even in countries with far stronger labor laws than the United States has. And state and local governments, the primary employers of government workers, will remain under pressure from businesses that can easily be bribed to move to other states and from Republican legislators determined to outsource and contract out every possible government function to the private sector.

Many of the trends forcing down the share of national income going to labor and hence the middle class are deep and structural, not necessarily caused by any specific set of governmental policies. But it doesn’t follow that governments can do nothing about them; it just requires analysis, creativity, and a willingness to use government and not denigrate it. The first step is to recognize that we have a problem that the free market is not going to fix.