The Auditing Roadblock: It’s Not Just China

The Public Company Accounting Oversight Board is out with a list of 58 international audit firms that it has been unable to inspect for at least four years. Those firms all audited companies registered in the United States, and â€" under American law â€" must be inspected by the board. The board has tried to arrange joint inspections with regulators from other countries, but a variety of impediments have arisen.

China, which is forever talking about cooperation but somehow never actually providing much of it, leads the list. There are eight Chinese firms and another eight based in Hong Kong. In some cases, the board says, it has been able to inspect Hong Kong firms, either because they did not audit mainland companies or because no objections were raised even though they did. In others, the Chinese authorities intervened to prevent inspections.

But for sheer numbers, Europe is the biggest scope of the problem. Forty of the audit firms are based in the European Union, where each country can decide for itself if it wants to be cooperative, and some are distinctly more eager than others to do so.

Italy, France and Sweden lead the European list with five uninspected audit firms each, followed by Belgium with four. There are none in Britain, and only one in Germany. The other countries with at least one uninspected audit firm are Austria, Cyprus, Czech Republic, Denmark, Greece. Hungary, Ireland, Luxembourg, Poland, Portugal and Spain.

The two uninspected firms in Latin America are both in Venezuela.

Most of the firms are members of big international groups of accounting firms. That means it is quite likely that some of them participated in audits of American internationals that had operations in the countries. The American firm may have signed the audit, but it was relying on the uninspected affiliate for what could have been critical information.

Based on their names, PricewaterhouseCoopers leads with 12, followed by Deloitte with 10, KPMG with nine, Ernst and BDO with eight each, Grant Thornton with three and Mazars, which is headquartered in France, with two. There are another six firms whose names do not indicate such an affiliation.

There is no way from the outside to know if an audit firm is doing its job, which is why inspections are crucial. Since the oversight board was established in 2002, it has found a lot of sloppiness and failures to perform basic tests, among American firms. There is some evidence those firms are getting better as a result. The board has found small American firms that were certifying Chinese audits that they basically did not do. And Chinese affiliates of the major firms have audited, and approved, financial statements of Chinese companies that turned out to be fraudulent. That does not prove their audits were badly done â€" good audits can miss frauds â€" but it does provide an indication that improvements may be necessary.

The Employer Mandate: Dukakis All Over Again

Casey B. Mulligan is an economics professor at the University of Chicago. He is the author of “The Redistribution Recession: How Labor Market Distortions Contracted the Economy.”

Last week’s adjustment to the employer mandate represents another battle in the political war of attrition between employers and those who want to carry out the federal health reform law. A similar war was fought in Massachusetts in the years after the Michael Dukakis administration and continued for decades.

In 1988, Governor Dukakis pushed through a law (not to be confused with the health law signed by Gov. Mitt Romney almost two decades later) that sought to achieve universal health insurance coverage in Massachusetts with a legislative package that included a $1,680 penalty per employee per year on employers who did not provide health coverage to their employees.

The Dukakis package passed by a narrow margin; to get it through the legislature, the law provided for a 45-month delay before the employer mandate took effect. (I recommend a book and a paper by John E. McDonough chronicling his health reform efforts as a Massachusetts legislator.)

Forty-four months after passage, the Massachusetts mandate was delayed three more years by a new law passed by the legislature over Gov. William Weld’s veto. The employer mandate was delayed twice more. As the fourth date approached for carrying out the 1988 employer mandate, the legislature repealed it entirely.

The employer mandate would return again in 2006 under Governor Romney, with its penalty set at one-tenth that of the Dukakis law.

These events are shown in the timeline below. Red shapes are dates of legislation and blue shapes are dates for putting the legislation into effect. The Massachusetts timeline is on the left, and that for the federal Affordable Care Act, which was passed in March 2010, on the right. (For both laws, month zero is when the law was passed). The timeline for the Affordable Care Act has a number of parallels with that of the Dukakis law. Both laws originally provided for an employer mandate delay: 45 months in Massachusetts and 46 months for the federal law.

In both cases, the employer mandate was delayed as the original effective date approached and businesses complained. The first federal delay was announced in the 40th month subsequent to the original law; the first Massachusetts delay was enacted in the 44th month.

In both cases, a second delay was announced, but it did not happen in Massachusetts until Month 80. Last week, already in Month 47 of the law, the federal government announced a second delay of the full employer mandate until January 2016 (which will be Month 70). So far, it looks as though a few employers will pay a reduced penalty as early as next January.

Both laws were passed when Democrats held the executive position and a majority in the legislature. In both cases, one-party rule ended soon after the health law was signed.

In Massachusetts, the legislature took the initiative to delay the employer mandate and carried out its decision as new laws. Federally, the Obama administration made the delays by notifying employers that the penalty either would not be enforced or would be enforced to a lesser degree than the original law specified..

The penalty amounts in the two laws are also similar. Adjusted for inflation, the Dukakis law’s penalty (federal tax deductible) was equivalent to a $2,791 penalty in 2014. The federal law’s penalty in 2014 would have effectively been $3,046 because it is not deductible from an employer’s business taxes.

One hundred months after the Dukakis law was passed, its employer penalty was finally repealed. If the federal law were to follow the same timeline, it would be repealed in June 2018, in the second year of a new presidency.

The Impact of a Minimum-Wage Increase

Jared Bernstein is a senior fellow at the Center on Budget and Policy Priorities in Washington and a former chief economist to Vice President Joseph R. Biden Jr.

The Congressional Budget Office has just released an analysis of the impact of increasing the federal minimum wage.  The budget office examined two proposals: an increase from its current level of $7.25 an hour to $9, and an increase in three annual steps of $0.95, reaching $10.10 in 2016.  Since the increase to $10.10 is the proposal supported by the White House and many congressional Democrats, that’s the one I’ll focus on here.

The most important finding is that on balance, low- and moderate-income Americans are big winners from a higher minimum wage, which would raise earnings and incomes, lower poverty and inequality, and do so at no net cost to the federal budget.

These are among the report’s key findings:

It estimates that 16.5 million low-wage workers would directly benefit from the proposed increase to $10.10 by the second half of 2016.

It further notes that because of “spillover effects” â€" the fact that employers typically increase the wages of workers slightly above the new minimum (the report estimates that the spillover will go up to $11.50) â€" an additional eight million low-wage workers are also likely to receive some benefit from the change.  

It estimates that the increase in the wage would reduce employment by about 500,000.  That amounts to about 0.3 percent of total employment, and about 3 percent of directly affected workers (500,000 of 16.5 million) and 1.5 percent of the total, including spillovers (i.e., including all those earning up to $11.50).

While those against the increase will highlight this employment loss finding as a rationale for their opposition, it is in fact entirely consistent with the view of most supporters of the increase: while the increase is expected to cause some job losses, the number of workers who would get a raise far outweigh those displaced: 97 percent to 98.5 percent of potentially affected workers would benefit from the proposal.

The budget office estimates that because of the increase, 900,000 who are currently poor would move above the poverty threshold.  That’s about 2 percent of the number it expects to be poor when the increase is phased in (45 million).

Of the affected workers, 88 percent are adult (20 and older), 56 percent are female, and most work full time (i.e., 53 percent work 35 or more hours a week).

The incomes of most families with low-wage workers will increase under the proposal.  About 70 percent of low-wage workers live in families whose average incomes are projected to rise, from 2.8 percent for the poorest families to 0.4 percent for middle-income families.

The wealthiest families, however, lose under the proposal, as their income is predicted to fall by 0.4 percent, or $700 for a family with average income around $180,000.  This results from reduced profits for business owners and slightly increased prices of goods and services that are not offset for these families by the higher wage.  Note that these income predictions imply slightly lower income inequality.

 
A few comments about these findings:

To derive the job-loss effects, the report does not do any original research.  It just uses estimates from a wide range of studies on the impact of past minimum-wage increases.  (For the technically inclined, it applies a negative employment elasticity that reflects the percent decline in jobs given a percent increase in the minimum wage.)  It is important to recognize that there is a very wide range of estimates from which the budget agency can choose, as shown in the chart below, which plots results of the employment effect from dozens of studies (from a recent set of slides from the White House Council of Economic Advisers).  This wide range does not imply that the budget office made a mistake, though it looks to me as if it applied a higher job-loss estimate than is the current consensus among economists who’ve closely studied the issue.

Note: “se” refers to standard error; 1/se is a measure of statistical significance. Source: Doucouliagos and Stanley (2009); data provided by John Schmitt.

As the chart shows, the employment impact from this “meta-analysis” clumps around zero, which is why the report finds that the policy is a significant net plus from the perspective of low-wage workers: Many more workers get a raise from the policy than are displaced from their jobs.

In fact, the study points out that the range, or confidence interval, around their central estimate ranges from a “very slight decrease” to one million.  The authors guess that there’s a two-thirds chance that the true estimate is in that range.

There is no policy I can think of that generates only benefits without any costs, and policy makers always have to weigh the two sides. In the case of the minimum wage, on the benefits side of ledger, the budget office shows that 16.5 million low-wage workers would directly get a much-needed pay increase at no cost to the federal budget.  Though the budget agency did not analyze longer-term results for these workers, it’s also the case that when those displaced by the increase get their next low-wage job, they too will benefit from a higher paycheck than would otherwise be the case.

As I’ve stressed many times on this blog, policy makers need to be concerned about the quantity of jobs, and pursue policies that will increase that number.  But they also have to worry about job quality, especially in the low-wage sector, where the decline in the real value of the minimum wage, the increase in earnings inequality (meaning less growth finds its way to the low end of the wage scale), and the low bargaining power of the work force have placed strong, negative pressure on wage trends for decades.

With such job-quality concerns in mind, I’d say the long history of research shows that increasing the minimum wage is a simple, effective policy that achieves its goal of raising the value of low-wage work with minimal distortions at no cost to the federal budget.  The Congressional Budget Office report further confirms that conclusion.

The Bank Rescue, Five Years Later

Matthew Cavanaugh/European Pressphoto Agency Timothy F. Geithner walking off stage after announcing new measures to stabilize the nation’s banks on Feb. 10, 2009.

Phillip Swagel is a professor at the School of Public Policy at the University of Maryland and was assistant secretary for economic policy at the Treasury Department from 2006 to 2009.

Sometimes government programs that seem flawed at their launch turn out to succeed against all expectations. No, this is not a post about the Affordable Care Act â€" I still think that will prove to be an unsustainable fiscal train wreck. I have in mind the Obama administration’s Financial Stability Plan to continue the bank rescue, kicked off in a speech by Timothy F. Geithner as Treasury secretary just over five years ago, on Feb. 10, 2009.

Mr. Geithner sought to explain how the new administration would carry on with the job of stabilizing still-fragile financial markets.  The Bush Treasury, at which I was a senior official, had intervened to support money market funds, banks, General Motors and Chrysler, and the American International Group insurance company (for which TARP money was used to restructure the Federal Reserve’s loan). The Federal Reserve and Federal Deposit Insurance Corporation had taken a range of vital and innovative actions to stanch the crisis, and staff members from the Treasury and Fed were developing a program to address credit market strains that were hindering business and consumer lending.

Even though Mr. Geithner had made key contributions to the financial rescue programs while in charge of the Federal Reserve Bank of New York, it was natural to expect the recently inaugurated President Obama to put his own stamp on the program, especially when many Americans were understandably discomfited at the idea of the bank rescue in the first place.  Indeed, the president himself had built up expectations just a few days earlier that his Treasury chief would provide “a new strategy to get credit moving again.”

The Geithner plan aimed to assure investors of bank stability, to cleanse bank balance sheets of remaining illiquid assets such as subprime mortgage-backed securities, and to encourage new lending to businesses and consumers. In broad strokes, this had the makings of an appropriate response to the problems facing the financial system, in large part by continuing and building on the efforts already under way that had succeeded in arresting the panic of September 2008.

Americans have long come to expect the Obama administration to overpromise and underdeliver.  But the introduction of the financial stability plan took place before this pattern became clear, and market participants were dismayed that the secretary’s speech had little information on how the worthwhile goals it set out would be achieved.

Trust in the stability of banks gradually had been rebuilt after the October 2008 announcement of TARP capital injections.  The improvement after following the speech, as can be seen in Figure 15 of a report issued by the Treasury in September 2009, which shows an increased cost to buy insurance against bond defaults by major banks. (It is admirable that the Geithner Treasury marked the timing of the secretary’s speech on charts in its report even while knowing that this would highlight the self-inflicted wound.)

I remember the feeling of dismay at the seemingly half-baked plan while watching the speech with others at a New York City asset management firm where I was talking about the economy. The reaction of equity markets mirrored the unhappiness in that room, with stocks down more than 5 percent that day.

It turns out, however, that the program sketched out by Mr. Geithner both came to pass and made a difference. Five years later, the United States financial sector is in much better condition. Banks have absorbed losses from loans made during the bubble and rebuilt their capital, and investor confidence has returned. Mr. Geithner’s proposals did not all work right away or in the scale initially envisioned.  In the end, however, Secretary Geithner deserves credit for making good on what he promised.

By far the most important component of Mr. Geithner’s proposal was a stress test to assess whether banks had “sufficient financial strength to absorb losses and to remain strongly capitalized” in the event of another severe downturn.  Led by the Federal Reserve, the stress test involved coming up with forecasts of banks’ financial positions in the event of a hypothetical renewed recession and housing price collapse.  Banks that could not make it through this very negative economic scenario would be given six months to raise capital from private investors, after which regulators would have pressed them to accept more TARP capital.

It is hard to remember, but back in the spring of 2009 many people believed that the United States government would nationalize large banks, starting with Citigroup and then moving on to Bank of America and perhaps others. Indeed, according to the New Yorker writer Ryan Lizza, this was a well-founded concern in the sense that nationalization was discussed as a policy option by the Obama administration amid a swell of sentiment for this action among economists and columnists.

In reality, nationalization was unlikely from the start for both legal reasons and political ones, the latter including that administration officials did not trust the F.D.I.C. to handle a seized megabank, as would have been required by law.  But investors did not realize this, and thought that new equity investments in banks would be at risk of being wiped out by a government takeover.

Concern over the possibility of force majeure was heightened by the Obama administration’s erratic response in March 2009 to the political furor involving bonuses to A.I.G. executives, with President Obama ordering officials to find ways to block the bonuses even after his staff had already declared that this was not possible under the law. Together, these factors led to a halt to capital raising by banks in the first quarter of 2009 (as can be seen in Figure 1 on Page 7 of the September 2009 Treasury report).  Just when financial stability and economic recovery depended on a stronger banking system, investor fears â€" some reflecting the administration’s own missteps â€" stood in the way.

The stress test results announced in May 2009 dispelled these concerns.  Ten of the 19 banks examined were required to raise $75 billion of additional capital between them, but the needed amounts were widely seen as attainable for all but GMAC Bank (since renamed as Ally Bank and still having problems with the Fed’s stress tests). On the whole, the first stress test was seen as a credible indication that major American banks would not fail â€" and thus were not at risk of being taken over by the government.  Citigroup in the end converted government preferred sharesfrom TARP into common stock, substantially diluting the holdings of many pre-crisis investors.  But the company remained afloat.

For market participants, the stress test results were the equivalent of an “all clear” flag at the beach after a storm â€" it was safe to go back into the water of banking industry investments (and they did, as noted a year later by the Fed chairman, Ben S. Bernanke).  A straight line can be drawn between the 2009 stress tests and today’s stronger American banking system â€" not at all what was expected given the initial reaction to Secretary Geithner’s speech.

The initiative to support new consumer and business lending announced by Secretary Geithner was the Term Asset-Backed Securities Lending Facility (TALF) under development since the fall of 2008 â€" Steven Shafran, the brilliant Treasury official who had come up with the idea, stayed on into the new administration to get the program up and running. The TALF was designed to restore the flow of securitized lending for auto purchases, student loans, business equipment, and other activities in which individual loans were packaged together into securities that were then sold to investors. This lending had fallen off sharply since the crisis manifested in August 2007.  TALF was a complement to the Federal Reserve’s quantitative easing program in the sense that the Fedpurchases of Treasury bonds and mortgage-backed securities were meant to push down overall long-term interest rates while TALF was aimed at specific market strains.

The program had a clever design: it was attractive to use during times of market stress, when funding to make loans was expensive, but it became relatively costly and thus less desirable once markets returned to normal. New lending was closed in June 2010 with $43 billion in loans outstanding.  A full discussion of TALF is a topic for a different post (though still relevant in that there are proposals even within the Fed for what amounts to a permanent TALF).  But an evaluation of TALF lending found that the program resulted in lower interest rates for the target activities while not favoring any particular security.  In other words, TALF worked effectively to address a problem that hindered particular types of household and business spending.

The last piece of the Geithner plan was for a Public-Private Investment Program (PPIP) in which the Treasury put government money alongside private investors to buy up the so-called legacy bad assets â€" the illiquid mortgage-backed securities that had figured so prominently in the crisis.  It turned out that the additional capital raised after the stress tests was more important for restoring confidence in the financial system than government assistance in cleaning up the old bad assets.  In the end, PPIP was a modest program (though with a positive financial return for taxpayers), but the problem at which it was aimed was largely addressed elsewhere.

As Mr. Geithner put it in May 2009 after the release of the stress tests, he was looking for banks “to get back to the business of banking.” This is what has happened. To see the alternative, consider the situation in euro zone countries, where multiple rounds of stress tests have not convinced investors that banks are sound, and weak credit growth remains a problem for economic growth.

Mr. Geithner is writing a book and can eventually give his account of these and other initiatives while he was Treasury secretary, including those on housing, where the results are still widely seen as not living up to the administration’s promises. On the bank rescue, however, the passage of time casts a considerably more favorable light than could possibly have been imagined back when his financial stability plan was announced.

Taking Their Investments Elsewhere

The Treasury on Tuesday released its estimates of international capital flows into and out of the United States, and they indicate that foreigners were net sellers of $40.2 billion of American stocks in 2013. That is hardly a large amount, but it is the first year since 1992 that they were net sellers.

Historically, foreigners have been considered to be contrary indicators in almost every stock market around the world, on the theory that they are less familiar with the market than are local investors, and more likely to buy at the top, after prices have been bid up to levels that could be unsustainable, and to sell at the bottom, after the local investors have fled, sending prices down.

Source: United States Treasury, via Haver Analytics.

The two years with the largest inflows of foreign money into American stocks were 2000 and 2007 â€" both years that the market hit historic peaks and began large declines. That would seem to confirm the theory.

Since the figures began to be estimated in 1979, the only previous years for net sales were 1984, 1988, 1990 and 1992. The Standard & Poor’s 500-stock index rose in each year after those sales, providing more evidence that foreigners tend to have poor market timing. Those periods of net foreign selling tended to follow periods of stock market weakness, such as the 1987 stock market crash or the 1990 recession, which reasonably could have scared investors.

But the 2013 sales came during one of the strongest years for the stock market in years, with the S.&P. 500 rising nearly 30 percent. And the market had been strong in the previous year as well. So this selling did not reflect recent market weakness.

Perhaps the explanation lies in a broader context. During the 1980s and early 1990s, there was widespread pessimism regarding the ability of the United States to compete with Japan, the rising economic power of that era. That pessimism proved to be unfounded, but it made investors nervous.

In the current era, there is worry that the United States cannot compete with the Chinese, and to a lesser extent with other emerging Asian economies. Perhaps the 2013 selling reflected pessimism about the country’s future, coupled with concern over the country’s ability to govern itself. The heaviest selling came during the summer, amid talk of a government shutdown or default, and during the last two months of the year, after the partial shutdown occurred in October.

The Ebb and Flow of Executive Power

Bruce Bartlett held senior policy roles in the Reagan and George H.W. Bush administrations and served on the staffs of Representatives Jack Kemp and Ron Paul. He is the author of “The Benefit and the Burden: Tax Reform â€" Why We Need It and What It Will Take.”

Executive orders have been controversial since the founding of the republic. They have long given rise to complaints from members of Congress that they infringe on its legislative power. But at the same time, it is obvious that legislative language cannot contemplate all of the means by which the laws are to be enforced; some executive latitude is clearly necessary.

Conflicts between Congress and the White House on executive orders tend to be greatest when each is under control of a different party. I can find no evidence that the Republican-controlled Congresses of the 1920s objected to the 1,203 executive orders issued by Calvin Coolidge, nor did the Democratic Congresses of the 1930s complain about Franklin D. Roosevelt’s 3,522 executive orders.

President Obama has issued 168 executive orders and at this rate probably will issue fewer during his two terms than George W. Bush, who issued 291.

The modern history of congressional concern about executive overreach begins with Richard Nixon. His Executive Order 11615 on Aug. 15, 1971, was among the most expansive in history, freezing all wages and prices in order to control inflation.

While Nixon’s action was controversial, the controversy was mainly regarding the efficacy of wage and price controls to deal with inflation. There was little criticism, even from Democratic members of Congress, as to his legal authority.

What really got Congress upset about Nixon’s usurping its power was the almost forgotten issue of “impoundment.” This refers to the long-exercised presidential practice of not spending funds appropriated by Congress when the president deemed the expenditure to be unnecessary or ill advised.

For example, in his State of the Union address in 1803, Thomas Jefferson informed Congress that he would not spend $50,000 (about $1.7 billion today) it had appropriated for gunboats on the Mississippi, thinking them unnecessary. (Please note that here and in all subsequent conversions of dollar sums to comparable value today, I do not rely on the change in the consumer price index but rather on a calculation of value at the time relative to the total output of the economy. I have been guided in this by the methodologies proposed by the founders of Measuring Worth.)

According to a 1974 law review article by the legal scholar Nile Stanton, “Every president from George Washington to Richard Nixon has almost certainly impounded appropriated funds.”

Congressional complaints about Nixon’s impoundment of appropriated funds began in 1971, when he withheld $13 billion (about $180 billion today). This was not his first use of impoundment, but the size of this action got the attention of conservatives in Congress who normally supported cuts in domestic spending.

Senator Sam J. Ervin Jr., Democrat of North Carolina, said impoundment “poses a threat to our system of government and patently violates the separation of powers doctrine.”

The following year, Nixon angered many Republicans when he impounded $202 million ($2.6 billion today) for food stamps, 10 percent of its budget. Senator Clifford Case, Republican of New Jersey, was especially upset that the White House never announced the impoundment, and efforts to find out who actually ordered the cut were stymied by the Agriculture Department’s bureaucracy.

After the 1972 election, the impoundment controversy heated up. On Jan. 6, 1973, the Senate voted to demand an accounting of impounded funds no later than Feb. 5. On that day, the Office of Management and Budget issued a list of $8.7 billion ($100 billion today) in impoundments across more than a dozen departments and agencies. Democrats complained that the list understated the true amount of impoundments, which they pegged at $12.2 billion ($140 billion today).

In an editorial, The New York Times said that while it was true that previous presidents had impounded appropriated funds, they did so only in isolated and temporary circumstances. By contrast, Nixon was asserting this power “across the board and as a permanent weapon in his constitutional arsenal.”

In a series of articles beginning on March 4, 1973, The New York Times reported that growing numbers of constitutional scholars viewed Nixon’s actions as an unprecedented expansion of executive power. It quoted the political scientist Thomas E. Cronin as saying of Nixon, “He has systematically gone about trying to strengthen the presidency in a great number of ways, frequently by circumventing the Constitution or expanding on past practices that were ambiguous or questionable.”

It is possible that Nixon would have prevailed on the impoundment issue if the debate had been limited to the narrow question of whether the president could withhold funds appropriated by Congress. But the impoundment issue quickly became conflated with Watergate when that investigation blew open on March 23, as James W. McCord Jr., who had been convicted in the burglary of the Democratic headquarters in 1972, said he and other defendants had been under “political pressure” to plead guilty and remain silent.

From that point forward, the Watergate scandal enveloped the White House until Nixon was forced to resign the presidency on Aug. 8, 1974. One of Nixon’s last acts as president was to sign, on July 12, 1974, the Congressional Budget and Impoundment Control Act of 1974, which permanently stripped the president of impoundment authority. Under ordinary circumstances, he would almost certainly have vetoed it. But in his weakened political position, Nixon had little choice since it undoubtedly would have been overridden by Congress.

I will have more to say about the long-term consequences of the Congressional Budget Act of 1974 in a future post.

Another long-term effect on the presidency in the struggle over Watergate, impoundment and presidential power is the deep impact it had on many Republicans who lived through it. In particular, Donald H. Rumsfeld and Dick Cheney, both of whom served President Ford as White House chief of staff, came away believing that the presidency had been hobbled by Congress and the president had been denied powers rightfully given to him by the Constitution. Another Ford administration official who shared this view was Antonin Scalia, now a Supreme Court justice, who then headed the Office of Legal Counsel at the Justice Department.

When, under President George W. Bush, Mr. Cheney and Mr. Rumsfeld occupied high-level positions, as vice president and as secretary of defense, they consciously sought to restore the president’s power that had been surrendered by Nixon. They were highly successful, bequeathing to President Obama a significantly strengthened presidency, which Republicans now berate Obama for using.

Before Blaming the Robots, Let’s Get the Policy Right

Jared Bernstein is a senior fellow at the Center on Budget and Policy Priorities in Washington and a former chief economist to Vice President Joseph R. Biden Jr.

The economist Alan Blinder and I were recently discussing whether technology was making a serious dent in job growth. Technically, we were considering whether the pace at which labor-saving technology is entering the workforce has accelerated â€" that is, whether the likelihood of “technological unemployment” had grown â€" when he suggested this thought experiment:

Say you were Thomas Jefferson’s chief economist and you’d just somehow seen a report from the year 2013 showing that 1.5 percent of the workforce was in agriculture, as opposed to the 90 percent in your day. You ran to the president with news of this crisis, telling him we’ve got to start preparing for mass unemployment.

Alas, like most similar warnings throughout history, yours would have been wrong. While productivity in farming has grown tremendously, displacing millions of workers, they’ve mostly found work elsewhere. That’s not at all meant to dismiss the disruption caused by technological progress on the lives of those displaced. It is simply to state that the predictions of the type we’re hearing more and more of â€" these days, about how robotics and digitized intelligence are finally going to cast us into unemployment in large numbers â€" have been wrong for centuries.

Of course, that doesn’t mean they’re wrong now. And in fact, a few years ago I plotted some data points that surprised me and many others.  The plot simply compared the growth of productivity to that of employment (using “full-time equivalents” for the latter, so that two half-time jobs are counted as one full-time job).

Sources: Bureau of Economic Analysis and Bureau of Labor Statistics. Productivity is for nonfarm business sector; employment is for full-time equivalents (F.T.E.’s) in the private sector.  F.T.E. employment is available only through 2012, so the 2013 value is projected using private payroll employment and part-time employment for 2013.

The standard view, the one that dismisses Jefferson’s shocked economist, derives from the fact that the two lines track each other for decades. I’ll explain why that’s been true in a moment, but what’s surprising here is the fact that this decades-long relationship appears to have broken down in recent years. A natural question is whether that gap toward the end of the chart is the work of robots or other labor-displacing technologies of the type discussed by Erik Brynjolfsson and Andrew McAfee in their compelling new book on this question, “The Second Machine Age.”

Before we get to the split in the lines, let’s discuss the decades of uniform growth (and longer historical charts show the same thing). A tiny bit of math from a recent analysis by the economist Lawrence Mishel, a critic of the robots-are-coming hypothesis, turns out to be elucidating.

First, Y = (Y/L)*L where Y is output, or gross domestic product, and L is hours worked, which you can think of as jobs. Now, with a bit more algebra, including taking natural logs, you quickly end up with:

Job Growth = Output Growth â€" Productivity Growth

With constant output growth, more efficient production (faster productivity growth) means fewer jobs, which is pretty commonsense: If we can produce the same output in fewer hours, we need less work to hold steady. But of course output has been anything but constant. The intervening variable, which has always glued those two lines together, is greater demand for the additional goods and services we can produce by dint of our increasing productivity. The fact that more efficient production typically lowers prices is a key part of the demand boost.

Therefore, the first thing your average macroeconomist would think when looking toward the end of the figure above is not “robots!” It’s “weak growth!” And that, in fact, is what Mr. Mishel finds. Here are his annualized growth numbers to fill into the little formula above for the last two business cycles: 1989-2000 and 2000-2007 (of course, since 2007, both growth and jobs have been cyclically depressed). Remember, it’s job growth equals G.D.P. growth minus productivity growth.

The 1990s: 1.5% = 3.3% â€" 1.8%

The 2000s: 0.3% = 2.4% â€" 2.1%

Looking to the right of the equal sign, what really changed in the 2000s was output growth, which fell almost a point (to 2.4 percent from 3.3 percent) over the period where the lines above diverge. Yes, productivity accelerated a bit, but had growth remained constant at 3.3 percent, we would have added jobs at a rate of 1.2 percent per year (3.3 percent minus 2.1 percent) instead of a measly 0.3 percent. That fourfold difference would have amounted to seven million more full-time jobs.

So weak growth and demand is the culprit. But “weak demand” is itself a pretty amorphous explanation. Perhaps that maps onto the tech story, with capital equipment replacing workers faster than normal, so that there’s more unemployment, less consumer spending, and weaker output growth.

Maybe, but that, too, is hard to see in the numbers, and it also has some tough competitors. That is, there are other explanations for the especially weak demand growth in the 2000s. Capital investment â€" spending on business equipment, including software and robotics â€" slowed in the 2000s relative to the 1990s, and while productivity accelerated slightly, it too has since slowed. On the other hand, Moore’s Law means that as time progresses, businesses can buy a lot more computing power for a lot less.

Still, this line of reasoning hardly makes sense. If technological advances lead to both fewer jobs and less growth for the length of a whole business cycle, why undertake them?

In fact, the 2000s were characterized by particularly large trade deficits that sapped domestic demand, along with a large and destructive increase in financial “innovation” that fueled a housing bubble from which we’re still recovering. True, the bubble spun off a construction boom and “wealth effect” that created some jobs in those years, but not enough to offset the slowdown in underlying demand. And again, exploding synthetic derivatives and a housing glut are not what the technologists have in mind when they analyze how robotics and artificial intelligence are improving the economy. Since that bubble burst, joblessness has been driven by cyclical forces and austere public policy.

None of this is dispositive evidence against either President Jefferson’s economist or Mr. Brynjolfsson and Mr. McAfee, who present a lot of anecdotal indicators of labor-saving advances that could conceivably increase the rate of technological unemployment going forward. But for now, I see nothing in either the data or the anecdotes that leads me to conclude that labor-saving technology is preventing us from getting back to full employment. To explain that, I’d invoke a far more evident culprit: the damaging policy set that has brought us bubbles and busts coupled with political resistance to the necessary policies to help get things back on track.

My argument regarding the question posed above is simple: Before we can assess the state of technological unemployment, we have to get rid of bad policy-induced unemployment. At that point, perhaps we can get a better handle on what the robots are really up to.