February 22, 2016
Okay, that’s not exactly true, but the new Economic Report of the President (ERP) has an interesting section that provides insight into the question of how fast the economy can grow, and more importantly how low the unemployment rate can go. The ERP re-examined the evidence on the relationship between inflation and unemployment.
Economists have long held the view that lower rates of unemployment would be associated with rising rates of inflation and vice versa. When the Federal Reserve Board decides to raise interest rates to slow the economy it is based on the belief that unemployment is falling to a level that would be associated with a rising rate of inflation.
Most economists now put the unemployment rate at which inflation starts to rise somewhere near the current 4.9 percent rate. (This is called the non-accelerating inflation rate of unemployment or NAIRU.) So does the ERP. But its analysis suggests a somewhat different story.
Figure 2-xiv shows the estimate of the NAIRU based on a simple regression measuring the change in the inflation rate against the level of unemployment using data from the last twenty years. The graph shows that the estimate has been falling consistently over the last two decades and is now near 4.0 percent. Furthermore, because the relationship is so weak, there is a huge range of uncertainty around this estimate. In fact, the figure shows that a zero percent unemployment rate is within the 95 percent confidence interval. (Don’t try that at home folks.)
The other very important part of this story is shown in Figure 2-xv. This graph shows the estimate of the coefficient on the unemployment rate. This coefficient is the estimate of how much the inflation rate rises as a result of the unemployment rate being below the NAIRU. By my eyeballs, the most recent estimate of this number is around 0.1. This implies that if the unemployment were a full percentage point below the actual NAIRU for a whole year, then we should expect the inflation rate to rise by 0.1 percentage point.
In other words, suppose the actual is NAIRU were 4.9 percent, as many still believe. Suppose we pushed the unemployment rate down to 3.9 percent for a full year. The Council of Economic Advisers’ regressions imply that the inflation rate would then rise by 0.1 percentage point. This means that if we starting with an inflation rate of 1.6 percent (roughly the current pace), then we will see the inflation rate rise to 1.7 percent.
This all matters for the Sanders’ growth projections, because it raises the issue of how far we should be looking to press the economy. When the Fed raised interest rates back in December, it was indicating its concern that the economy was growing too fast and that unemployment could get too low.
The point of raising interest rates was to reduce the rate of job creation. The quarter point rate hike in December did not have much impact, but a series of future rate hikes will have us paying more car loans and mortgages, and businesses paying more to borrow for new investment, and state and local governments paying higher interest rates on bonds to finance infrastructure. This will slow the economy and keep people from getting jobs.
However if we look at the research in the ERP, it suggests both that we are still far from the NAIRU and that there would be very little consequence from going below the NAIRU for relatively brief periods of time. This should be a very strong argument for pushing the economy further, as Senator Sanders has proposed to do with his aggressive program.
That is hardly an endorsement of the specifics or the even the size of the Sanders agenda (and certainly not the now famous growth projections from Gerald Friedman), but it does argue for pushing the envelope in terms of bringing down the unemployment rate. As Jared Bernstein and I argued in our book a few years ago, low unemployment is probably the best way to help disadvantaged groups in society.
As a rule of thumb, a one percentage point drop in the unemployment rate for whites corresponds to a two percentage point drop in the unemployment rate for African Americans and a six percentage point drop in the unemployment rate for African American teens. There is a similar story for Hispanics. Disproportionately, the people hired will be those with less education.
Furthermore, we find that not only are less-educated workers more likely to get jobs, many will be able to get more hours on the job. This means that many people who are only working part-time will be able to get full-time employment. And perhaps most importantly, workers at the middle and bottom of the wage distribution will gain enough bargaining power to be able to get wage increases. The low unemployment years at the end of the 1990s were the only period in the last four decades in which workers at all points along the wage distribution were able to see sustained gains in real wages.
So we should look at the evidence in the ERP as making a very strong case for more government spending, especially in areas that will provide lasting benefits. The failure to have this spending means we are just throwing potential output in the toilet and needlessly keeping people from having jobs. Just think, the kids in Flint got poisoned because we didn’t feel like letting people have jobs. Sounds like a good reason, right?
Anyhow, we should want to see Congress approve more money to address the country’s needs, but even more immediately we should not want the Fed to go in the opposite direction. The last thing we need are further rate hikes from the Fed for the purpose of slowing the economy. This is the point of the Fed Up campaign sponsored by the Center for Popular Democracy. Urge a presidential candidate near you to sign on. (Senator Sanders is already on record criticizing the Fed’s rate hike.)
Comments