The Atlanta Fed's macroblog provides commentary and analysis on economic topics including monetary policy, macroeconomic developments, inflation, labor economics, and financial issues.
- BLS Handbook of Methods
- Bureau of Economic Analysis
- Bureau of Labor Statistics
- Congressional Budget Office
- Economic Data - FRED® II, St. Louis Fed
- Office of Management and Budget
- Statistics: Releases and Historical Data, Board of Governors
- U.S. Census Bureau Economic Programs
- White House Economic Statistics Briefing Room
June 01, 2018
Part-Time Workers Are Less Likely to Get a Pay Raise
A recent FEDS Notes article summarized some interesting findings from the Board of Governors' 2017 Survey of Household Economics and Decisionmaking. One set of responses that caught my eye explored the connection between part-time employment and pay raises. The report estimates that about 70 percent of people working part-time did not get a pay increase over the past year (their pay stayed the same or went down). In contrast, only about 40 percent of full-time workers had no increase in pay.
This pattern is broadly consistent with what we see in the Atlanta Fed's Wage Growth Tracker data. As the following chart indicates, the population of part-time workers (who were also employed a year earlier) is generally less likely to get an increase in the hourly rate of pay than their full-time counterparts. Median wage growth for part-time workers has been lower than for full-time workers since 1998.
This wage growth premium for full-time work is partly accounted for by the fact that the typical part-time and full-time worker are different along several dimensions. For example, a part-time worker is more likely to have a relatively low-skilled job, and wage growth tends to be lower for workers in low-skilled jobs.
As the chart shows, the wage growth gap widened considerably in the wake of the Great Recession. The share of workers who are in part-time jobs because of slack business conditions increased across industries and occupation skill levels, and median part-time wage growth ground to a halt.
While part-time wage growth has improved since then, the wage growth gap is still larger than it used to be. This larger gap appears to be attributable to a rise in the share of part-time employment in low-skilled jobs since the recession. In particular, relative to 2007, the share of part-time workers in the Wage Growth Tracker data in low-skilled jobs has increased by about 3 percentage points, whereas the share of full-time workers in low-skilled jobs has remained essentially unchanged. Note that what is happening here is that more part-time jobs are low skilled than before, and not the other way around. Low-skilled jobs are about as likely to be part-time now as they were before the recession.
How does this shift affect an assessment of the overall tightness of today's labor market? Looking at the chart, the answer is probably “not much.” As measured by the Wage Growth Tracker, median wage growth for both full-time and part-time workers has not been accelerating recently. If the labor market were very tight, then this is not what we would expect to see. The modest rise in average hourly earnings in the June 1 labor report for May 2018 to 2.7 percent year over year, even as the unemployment rate declined to an 18-year low, seems consistent with that view. A reading on the Wage Growth Tracker for May should be available in about a week.
May 31, 2018
Learning about an ML-Driven Economy
Developments in artificial intelligence (AI) and machine learning (ML) have drawn considerable attention from both the real and financial sides of the economy. The Atlanta Fed's recent Financial Markets Conference, Machines Learning Finance: Will They Change the Game?, explored the implications of AI/ML for the financial system and public policy. The conference also included two macroeconomics-related sessions. A presentation of an academic paper, and the subsequent discussion, looked at why AI/ML has not (yet) shown up in the productivity statistics. Also, a policy panel on the implications of AI/ML developments for monetary policy was part of the conference. This post summarizes the policy panel discussion.
Vincent Reinhart, chief economist at Standish Mellon Asset Management, opened the panel discussion with the observation that developments in AI/ML could affect the performance of the overall economy in a variety of ways. For example, advancing technology could better match workers with jobs and, as a result, boost employment. On the other hand, it could also complicate job matching by forcing jobs and workers to become more specialized.
A combination of three factors is driving the recent growth in AI/ML, explained Carolyn Evans, head economist and senior data scientist at Intel Corporation: increased data availability, faster computers, and improved algorithms for analyzing the data. Like Reinhart, she noted that AI/ML could have various effects on the economy. For example, AI/ML is helping to reduce cost and boost supply. On the demand side, AI/ML is increasing the efficiency of product searches by buyers. However, as some online sellers become better than others at using AI/ML to help customers find the products they want, customer relationships may become stickier. In addition, firms may come to value interactions with customers more highly because these interactions could provide them with valuable data to use with AI/ML to better serve current and future customers. Evans raised the question of whether these developments could change the nature of pricing.
Dallas Fed president Rob Kaplan said he believes AI/ML is causing a structural change. It is not the first new technology to affect the economy, but the economic effects of this technology are more pervasive. For instance, business pricing power is already more constrained than it used to be, but even businesses that seemingly have some power currently worry that they make themselves more vulnerable to AI/ML-enabled disruption if they raise prices. Kaplan also emphasized the importance of skills training and building human capital to alleviate what he views as the inevitable loss of jobs to AI/ML.
The issue of how monetary policymakers should think about AI/ML was the focus of a presentation by Chicago Fed president Charles Evans. He observed that the "sign, magnitude, and timing" of any resulting structural change are all uncertain. This uncertainty, he said, argues against the use of fixed policy rules such as the Taylor Rule. He suggested that the Federal Reserve should instead follow an "outcome-based policy," adjusting policy based on the evolution of expected inflation and unemployment relative to the policy objectives of stable prices and full employment.
You can download the available presentations from the 2018 Financial Markets Conference web pages. The videos will be posted as they become available. Read Notes from the Vault for a summary of sessions on the strengths and weaknesses of ML, some financial regulatory and broader ethical issues, and the use of ML by investors.
April 18, 2018
Hitting a Cyclical High: The Wage Growth Premium from Changing Jobs
The Atlanta Fed's Wage Growth Tracker rose 3.3 percent in March. While this increase is up from 2.9 percent in February, the 12-month average remained at 3.2 percent, a bit lower than the 3.5 percent average we observed a year earlier. The absence of upward momentum in the overall Tracker may be a signal that the labor market still has some head room, as suggested by participants at the last Federal Open market Committee (FOMC) meeting, who noted this in the meeting:
Regarding wage growth at the national level, several participants noted a modest increase, but most still described the pace of wage gains as moderate; a few participants cited this fact as suggesting that there was room for the labor market to strengthen somewhat further.
Although wages haven't been rising faster for the median individual, they have been for those who switch jobs. This distinction is important because the wage growth of job-switchers tends to be a better cyclical indicator than overall wage growth. In particular, the median wage growth of people who change industry or occupation tends to rise more rapidly as the labor market tightens. To illustrate, the orange line in the following chart shows the median 12-month wage growth for workers in the Wage Growth Tracker data who change industry (across manufacturing, construction, retail, etc.), and the green line depicts the wage growth of those who remained in the same industry.
As the chart indicates, changing industry when unemployment is high tends to result in a wage growth penalty relative to those who remain employed in the same industry. But when the unemployment rate is low, voluntary quits rise and workers who change industries tend to experience higher wage growth than those who stay.
Currently, the wage growth premium associated with switching employment to a different industry is around 1.5 percentage points and growing. For those who are tempted to infer that the softness in the Wage Growth Tracker might signal an impending labor market slowdown, the wage growth performance for those changing jobs suggests the opposite: the labor market is continuing to gradually tighten.
April 02, 2018
Thoughts on a Long-Run Monetary Policy Framework, Part 4: Flexible Price-Level Targeting in the Big Picture
In the second post of this series, I enumerated several alternative monetary policy frameworks. Each is motivated by a recognition that the Federal Open Market Committee (FOMC) is likely to confront future scenarios where the effective lower bound on policy rates comes into play. Given such a possibility, it is important to consider the robustness of the framework.
My previous macroblog posts have focused on one of these frameworks: price-level targeting of a particular sort. As I hinted in the part 3 post, I view the specific framework I have in mind as a complement to, and not a substitute for, many of the other proposals that are likely to be considered. In this final post on the topic, I want to expand on that thought, considering in turn the options listed in part 2.
- Raising the FOMC's longer-run inflation target
The framework I described in part 3 was constructed to be consistent with the FOMC's current long-run objective of 2 percent inflation. But nothing in the structure of the plan I discussed would bind the Committee to the 2 percent objective. Obviously, a price-level target line can be constructed for any path that policymakers choose. The key is to have such a target and coherently manage monetary policy so that it achieves that target. The slope of the price-level path—that is, the underlying long-run inflation rate—is an entirely separate issue.
- Maintaining the 2 percent longer-run inflation target and policy framework more or less as is, relying on unconventional tools when needed
As noted, the flexible price-level targeting example I discussed in part 3 was constructed with a long-run 2 percent inflation rate as the key benchmark. In that regard, it is clearly consistent with the Fed's current inflation goal.
Further, a central question in the current framework is how to interpret a goal of 2 percent inflation in the longer run. One interpretation is that the central bank aims to deliver an inflation rate that averages 2 percent over some period of time. Another interpretation is that the central bank aims to deliver an inflation rate that tends toward 2 percent, letting bygones be bygones in the event that realized inflation rates deviate from 2 percent.
The bounded price-level targets I have presented do not force a particular answer to the question I raise, and both views can be supported within the framework. Hence, the framework is consistent with whichever view the FOMC might adopt. The only caveat is that deviations from 2 percent cannot be so large and persistent that they push the price level outside the target bounds.As to the problem of the federal funds rate falling to a level that makes further cuts infeasible, nothing in the notion of a price-level target rules out (or demands) any particular policy tool. If anything, bounded price-level targets could expand the existing toolkit. They certainly do not constrain it.
Targeting nominal gross domestic product (GDP) growth
Targeting nominal GDP growth, which is the sum of real GDP growth and the inflation rate, represents a deviation from the price-level targeting I have described. In this framework, the longer-run rate of inflation depends on the longer-run rate of real GDP growth.
To see how this works, consider the period from 2003 to 2013. In 2003, the Congressional Budget Office projected an average annual potential GDP growth rate of 2.9 percent over the next 10 years. Had there been a nominal GDP growth target of 5 percent at this time, the implicit annualized inflation target would have been just over 2 percent. However, current CBO estimates indicate that actual potential GDP growth over this period averaged just 1.5 percent, which would suggest an inflation target of 3.5 percent. As data came in and policymakers saw this lower level of growth, they would have responded by shifting upward the implicit inflation target.
For advocates of using a nominal GDP target, shifting inflation targets is a key feature and not a bug, as it allows policy to adjust in real time to unforeseen cyclical and structural developments. What nominal GDP targeting doesn't satisfy is the principle of bounded nominal uncertainty. Eventually, price-level bounds that are set with an assumed potential real growth path will be violated if shifts in potential growth are sufficiently large. The appeal of nominal GDP targeting depends on how one weighs the benefits of inflation-target flexibility against the costs of price-level uncertainty inherent in that framework.
- Adopting flexible inflation targets that are adjusted based on economic conditions
Recently, my colleague Eric Rosengren, president of the Boston Fed, offered a proposal (here and here) that has some of the flavor of nominal GDP targeting but differs in important respects. Like nominal GDP targeting, President Rosengren's framework would adjust the target inflation rate given structural shifts in the economy. However, if I understand his idea correctly, the FOMC would deliberate specifically on the desired rate of inflation and adjust the target within a predetermined range.
Relying on the target's appropriate range opens the possibility of compatibility between President Rosengren's framework and the one I presented. Policymakers could use price-level targeting concepts in developing a range of policy options given the state of the economy. The breadth of the range of options would depend on the bounds the FOMC felt represented an acceptable degree of price-level uncertainty.
Summing all of this up, then—to me, the important characteristic of a sound monetary policy framework is that it provides a credible nominal anchor while maintaining flexibility to address changing circumstances. I think some form of flexible price-level targeting can be a part of such a framework. I look forward to a robust and constructive debate.
March 28, 2018
Thoughts on a Long-Run Monetary Policy Framework, Part 3: An Example of Flexible Price-Level Targeting
I want to start my discussion in this post with two points I made in the previous two macroblog posts (here and here). First, I think a commitment to delivering a relatively predictable price-level path is a desirable feature of a well-constructed monetary framework. Price stability is in my view achieved if people can have confidence that the purchasing power of the dollars they hold today will fall within a certain range at any date in the future.
My second point was that, as a matter of fact, the Federal Open Market Committee (FOMC) delivered on this definition of price stability during the years 1995–2012. (The FOMC formally adopted its 2 percent long-run inflation target in 2012.)
If you are reading this blog, you're almost certainly aware that since 2012, the actual personal consumption expenditures (PCE) inflation rate has persistently fallen short of the 2 percent goal. That, of course, means that the price level has fallen increasingly short of a reference 2 percent path, as shown in chart 1 below.
Is this deviation from the price-level path a problem? The practical answer to that question will depend on how my proposed definition of price stability is implemented.
By way of example, let's suppose that the FOMC commits to conducting monetary policy in such a way that the price level will always fall within plus-or-minus 5 percent of the long-run target path (which itself we define as the path implied by a constant 2 percent inflation rate). This policy—and how it relates to the actual path of PCE price inflation—is illustrated in chart 2.
So would inflation falling short of the 2 percent longer-run goal be a problem if the Fed was operating within the framework depicted in chart 2? In a sense, the answer is no. The current price level would be within the bounds of a hypothetical commitment made in 1995. If the central bank could perpetually deliver 2 percent annual inflation, that promise would remain intact, as shown in chart 3.
Of course, chart 3 depicts a forward path for prices whose margin for error is quite slim. Continued inflation below 2 percent would, in short order, push the price level below the lower bound, likely requiring a relatively accommodative monetary policy stance—that is, if policymakers sought to satisfy a commitment to this framework's definition of price stability.
Central bankers in risk management mode might opt for policies designed to deliberately move the price level toward the 2 percent average inflation midpoint in cases where the price level moves too close for the Committee's comfort to one of the bounds (as, perhaps, in chart 3). It bears noting that in such cases there are a wide range of options available to policymakers with respect to the timing and pace of that adjustment.
This scenario illustrates the flexibility of the price-level targeting framework I'm describing. I think it's important to think in terms of gradual adjustments that don't risk whipsawing the economy or force the central bank to be overly precise in its short-run influence on inflation and economic activity. A key feature of such a policy framework includes considerable short- and medium-run flexibility in inflation outcomes.
But the other key feature is that the framework limits that same flexibility—that is, it satisfies the principle of bounded nominal uncertainty. Suppose you and another person agree that you will receive a $1 payment in 10 years in exchange for a service provided today. If the inflation rate over this 10-year period is exactly 2 percent per year, then the real value of that dollar in goods and services would be 82 cents.
In my example (the one with a plus-or-minus 5 percent bound on the price level), monetary policymakers have essentially committed that the agreed-upon payment would not result in real purchasing power of less than 78 cents (and the payer could be confident that the real purchasing power relinquished would not be more than 86 cents).
The crux of my argument is that a "good" monetary policy framework limits the degree of uncertainty associated with contracts involving transfers of dollars over time. In limiting uncertainty, monetary policy contributes to economic efficiency.
The 5 percent bound I chose for my illustration is obviously arbitrary. The magnitude of the acceptable deviations from the price-level path would be a policy decision. I'm not sure we know a whole lot about what range of deviations from an expected price path contributes most consistently to economic efficiency. A benefit of the framework I am describing is that it would focus research, discussion, and debate squarely on that question.
This series of posts is going on hiatus for a few days. Tomorrow, the Atlanta Fed is going to release its 2017 Annual Report, and I certainly don't want to steal its thunder. And Friday, of course, will begin the Easter weekend for many people.
But I want to conclude this post by emphasizing that the framework I am describing is more of a refinement of, and not a competitor to, many of the framework proposals I discussed in Monday's post. This is an important point and one that I will turn to in the final installment of this series, to be published next Monday.
March 27, 2018
Thoughts on a Long-Run Monetary Policy Framework, Part 2: The Principle of Bounded Nominal Uncertainty
In yesterday's macroblog post, I discussed one of the central monetary policy questions of the day: Is the possibility of hitting the lower bound on policy rates likely to be an issue for the Fed going forward, do we care, and—if we do—what can we do about it?
The answers to the first questions are, in my opinion, yes and yes. That's the easy part. The last question—what can we do about it?—is the hard part. In the end, this is a question about the framework for conducting monetary policy. The menu of options includes:
- Raising the Federal Open Market Committee's (FOMC) longer-run inflation target;
- Maintaining the current policy framework, including the 2 percent longer-run inflation target, relying on unconventional tools when needed;
- Targeting the growth rate of nominal gross domestic product;
- Adopting an inflation range with flexible inflation targets that are adjusted based on the state of the economy (a relatively recent entry to the list suggested by Boston Fed president Eric Rosengren );
- Price-level targeting.
Chicago Fed president Charles Evans, San Francisco Fed president John Williams, and former Federal Reserve chairman Ben Bernanke, among others, have advocated for some version of the last item on this list of options. I am going to add myself to the list of people sympathetic to a policy framework that has a form of price-level targeting at its center.
I'll explain my sympathies by discussing principles that are central to my thinking.
First, I think the Fed's commitment to the long-run 2 percent inflation objective has served the country well. I recognize that the word “commitment” in that sentence might be more important than the specific 2 percent target value. But credibility and commitment imply objectives that, though not immutable, rarely change—and then only with a clear consensus on a better course. With respect to changing the 2 percent objective as a longer-run goal, my feet are not set in concrete, but they are in pretty thick mud.
Second, former Fed chairman Alan Greenspan offered a well-known definition of what it means for a central bank to succeed on a charge to deliver price stability. Paraphrasing, Chairman Greenspan suggested that the goal of price stability is met when households and business ignore inflation when making key economic decisions that affect their financial futures.
I agree with the Greenspan definition, and I believe that the 2 percent inflation objective has helped us meet that criterion. But I don't think we have met the Greenspan definition of price stability solely because 2 percent is a sufficiently low rate of inflation. I think it is also critical that deviations of prices away from a path implied by an average inflation rate of 2 percent have, in the United States, been relatively small.
Here's how I see it: until recently, the 2 percent inflation objective in the United States has essentially functioned as a price-level target centered on a 2 percent growth path. The orange line in the chart below shows what a price-level path of 2 percent growth would have been over the period from 1995 to 2012. I chose to begin with 1995 because it arguably began the Fed's era of inflation targeting. Why does the chart end in 2012? I'll get to that tomorrow, when I lay out a specific hypothetical plan.
The green line in the chart is the actual path of the price level, as measured by the price index for personal consumption expenditures. The chart explains what I mean when I say the FOMC effectively delivered on a 2 percent price-level target. Over the period depicted in this chart, the price level did not deviate much from the 2 percent path.
I believe the inflation outcome apparent in the chart is highly desirable. Why? Because the resulting price-level path satisfies what I will call the “principle of bounded nominal uncertainty.” In essence, the principle of bounded nominal uncertainty means that if you save a dollar today you can be “reasonably confident” about what the real value of that saving will be in the future.
For example, suppose that in January 1995 you had socked away $1 in cash that you intended to spend exactly five years later. If you believed that the Fed was going to deliver an average annual inflation rate of 2 percent over this period, you'd expect that dollar to be worth about 90 cents in real purchasing power by January 2000. (Recall that cash depreciates at the rate of inflation—I didn't say this was the best way to save!)
In fact, because the price level's realized path over that time hewed very closely to the expected 2 percent growth path, the actual value of the dollar you saved would have been very close to the 90 cents you expected. And this, I think, epitomizes a reasonable definition of price stability. If you and I enter into a contract to exchange a dollar at some future date, we can confidently predict within some range that dollar's purchasing power. Good monetary policy, in my view, will satisfy the principle of bounded nominal uncertainty.
This is the starting point of my thinking about a useful monetary policy framework—and how I think about price-level targeting generally. Tomorrow, I will expand on this thought and offer a specific example of how a price-level target might be put into operation in a way that is both flexible and respectful of the principle of bounded nominal uncertainty.
March 26, 2018
Thoughts on a Long-Run Monetary Policy Framework: Framing the Question
"Should the Fed stick with the 2 percent inflation target or rethink it?" This was the very good question posed in a special conference hosted by the Brookings Institution this past January. Over the course of roughly two decades prior to the global financial crisis, a consensus had formed among monetary-policy experts and practitioners the world over that something like 2 percent is an appropriate goal—maybe even the optimal goal—for central banks to pursue. So why reconsider that target now?
The answer to that question starts with another consensus that has emerged in the aftermath of the global financial crisis. In particular, there is now a widespread belief that, once monetary policy has fully normalized, the federal funds rate—the Federal Open Market Committee's (FOMC) reference policy rate—will settle significantly below historical norms.
Several of my colleagues have spoken cogently about this phenomenon, which is often cast in terms of concepts like r-star, the natural rate of interest, the equilibrium rate of interest, or (in the case of my colleague Jim Bullard ), r-dagger. I like to think in terms of the "neutral" rate of interest; that is, the level of the policy rate consistent with the FOMC meeting its longer-run goals of price stability and maximum sustainable growth. In other words, the level of the federal funds rate should be consistent with 2 percent inflation, the unemployment rate at its sustainable level, and real gross domestic product at its potential.
Estimates of the neutral policy rate are subject to imprecision and debate. But a reasonable notion can be gleaned from the range of projections for the long-run federal funds rate reported in the Summary of Economic Projections (SEP) released just after last week's FOMC meeting. According to the latest SEP, neutral would be in a range 2.3 to 3.0 percent.
For some historical context, in the latter half of the 1990s, as the 2 percent inflation consensus was solidifying, the neutral federal funds rate would have been pegged in a range of something like 4.0 to 5.0 percent, roughly 2 percentage points higher than the range considered to be neutral today.
The implication for monetary policy is clear. If interest rates settle at levels that are historically low, policymakers will have limited scope for cutting rates in the event of a significant economic downturn (or at least more limited scope than they had in the past). I think it's fair to say that even relatively modest downturns are likely to yield policy reactions that drive the federal funds rate to zero, as happened in the Great Recession.
My view is that the nontraditional tools deployed after December 2008, when the federal funds rate effectively fell to zero, were effective. But it is accurate to say that our experience with these tools is limited, and the effectiveness of those tools remains controversial. I join the opinion that, all else equal, it would be vastly preferable to conduct monetary policy through the time-tested approach of raising and lowering short-term policy rates, if such an approach is available.
This point is where the challenge to the 2 percent inflation target enters the picture. The neutral rate I have been describing is a nominal rate. It is roughly the sum of an inflation-adjusted real rate—determined by fundamental saving and investment decisions in the global economy—and the rate of inflation. The downward drift in the neutral rate I have been describing is attributable to a downward drift in the inflation-adjusted real rate. A great deal of research has documented this phenomenon, such as some influential research by San Francisco Fed president John Williams and Thomas Laubach, the head of the monetary division at the Fed's Board of Governors.
In the long run, a central bank cannot reliably control the real rate of interest. So if we accept the following premises...
- A neutral rate that is too low to give the central bank enough room to fight even run-of-the-mill downturns is problematic;
- Cutting rates is the optimal strategy for addressing downturns; and
- The real interest rate is beyond the control of the central bank in the long run
...then we must necessarily accept that raising the neutral rate, thus affording monetary policymakers the desired rate-cutting scope when needed, would require raising the long-run inflation rate. Hence the argument for rethinking the Fed's 2 percent inflation target.
But is that the only option? And is it the best option?
The answer to the first question is clearly no. The purpose of the Brookings Institution sessions is addressing the pros and cons of the different strategies for dealing with the low neutral rate problem, and I commend them to you. But in upcoming macroblog posts, I want to share some of my thoughts on the second question.
Tomorrow, I will review some of the proposed options and explain why I am attracted to one in particular: price-level targeting. On Wednesday, I will propose what I think is a potentially useful model for implementing a price-level targeting scheme in practice. I want to emphasize that these are preliminary thoughts, offered in the spirit of stimulating the conversation and debate. I welcome that conversation and debate and look forward to making my contribution to moving it forward.
March 23, 2018
What Are Businesses Saying about Tax Reform Now?
In a recent macroblog post, we shared some results of a joint national survey that is an ongoing collaboration between the Atlanta Fed, Nick Bloom of Stanford University, and Steve Davis of the University of Chicago, and Jose Barrero of Stanford University. (By the way, we're planning on calling this work the "Survey of Business Executives," or SBE.).
In mid-November, we posed this question to our panel of firms:
If passed in its current form, how would the Tax Cuts and Jobs Act affect your capital expenditures in 2018?
At the time, we (and perhaps others) were a little surprised to find that roughly two-thirds of respondents indicated that tax reform hasn't enticed them into changing their investment plans for 2018. Our initial interpretation was that the lack of an investment response by firms made it unlikely that we'd see a sharp acceleration in output growth in 2018.
Another interpretation of those results might be that firms were unwilling to speculate on how they'd respond to legislation that was not yet set in stone. Now that the ink has been dry on the bill for a while, we decided to ask again.
In our February survey—which was in the field from February 12 through February 23—we asked firms, "How has the recently enacted Tax Cuts and Jobs Act (TCJA) led you to revise your plans for capital expenditures in 2018?" The results shown below—restricted to the 218 firms that responded in both November 2017 and February 2018—suggest that, if anything, these firms have revised down their expectations for this year:
You may be thinking that perhaps firms had already set their capital expenditure plans for 2018, so asking about changes in firms' 2018 plans isn't too revealing—which is why we asked them about their 2019 plans as well. The results (showing all 272 responses in February) are not statistically different from their 2018 response. Roughly three-quarters of firms don't plan to change their capital expenditure plans in 2019 as a result of the TCJA:
These results contain some nuance. It seems that larger firms (those with more than 500 employees) responded more favorably to the tax reform. But it is still the case that the typical (or median) large firm has not revised its 2019 capex plans in response to tax changes.
Why the disparity between smaller and larger firms? We're not sure yet—but we have an inkling. In a separate survey we had in the field in February—the Business Inflation Expectations (BIE) survey—we asked Sixth District firms to identify their tax reporting structure and whether or not they expected to see a reduction in their tax bill as a result of the TCJA. Larger firms—which are more likely to be organized as C corporations—appear to be more sure of the TCJA's impact on their bottom lines. Conversely, smaller "pass-through" entities appear to be less certain of its impact, as shown here:
For now, we're sticking with our initial assessment that the potential for a sharp acceleration in near-term output growth is limited. However, there is some upside risk to that view if more pass-through entities start to see significantly smaller tax bills as a result of the TCJA.
March 06, 2018
A First Look at Employment
One Friday morning each month at 8:30 is always an exciting time here at the Atlanta Fed. Why, you might ask? Because that's when the U.S. Bureau of Labor Statistics (BLS) issues the newest employment and labor force statistics from the Employment Situation Summary. Just after the release, Atlanta Fed analysts compile a "first look" report based on the latest numbers. We have found this initial view to be a very useful glimpse into the broad health of the national labor market.
Because we find this report useful, we thought you might also find it of interest. To that end, we have added the Labor Report First Look tool to our website, and we'll strive to post updated data soon after the release of the BLS's Employment Situation Report. Our Labor Report First Look includes key data for the month and changes over time from both the payroll and household surveys, presented as tables and charts.
Additionally, we will also use the bureau's data to create other indicators included in the Labor Report First Look. For example, one of these is a depiction of changes in payroll employment by industry, in which we rank industry employment changes by average hourly pay levels. This tool allows us to see if payrolls are gaining or losing higher- or lower-paying jobs, as the following chart shows.
But wait, there's more! We will also report information on the so-called job finding rate—an estimate of the share of unemployed last month who are employed this month—and a broad measure of labor underutilization. Our underutilization concept is related to another statistic we created called Z-Pop, computed as the share of the population who are either unemployed or underemployed (working part-time hours but wanting full-time work) or who say they currently want a job but are not actively looking. We have found this to be a useful supplement to the BLS's employment-to-population ratio (see the chart).
The Labor Report First Look tool also allows you to dig a bit deeper into Atlanta Fed labor market analysis via links to our Human Capital Data & Tools (which includes the Wage Growth Tracker and Labor Force Dynamics web pages) and links to some of our blog posts on labor market developments and related research. (In fact, it's easy to stay informed of all Labor Report First Look updates by subscribing to our RSS feed or following the Atlanta Fed on Twitter.
We hope you'll look for the inaugural Labor Report First Look next Friday morning...we know you'll be as excited as we will!
February 28, 2018
Weighting the Wage Growth Tracker
The Atlanta Fed's Wage Growth Tracker (WGT) has shown its usefulness as an indicator of labor market conditions, producing a better-fitting Phillips curve than other measures of wage growth. So we were understandably surprised to see the WGT decline from 3.5 percent in 2016 to 3.2 percent in 2017, even as the unemployment rate moved lower from 4.9 to 4.4 percent.
This unexpected disconnect between the WGT and the unemployment rate naturally led us to wonder if it was a consequence of the way the WGT is constructed. Essentially, the WGT is the median of an unweighted sample of individual wage growth observations. This sample is quite large, but it does not perfectly represent the population of wage and salary earners.
Importantly, the WGT sample has too few young workers, because young workers are much more likely to be in and out of employment and hence less likely to have a wage observation in both the current and prior years. To examine the effect of this underrepresentation, we recomputed median wage growth after weighting the WGT sample to be consistent with the distribution of demographic and job characteristics of the workforce in each year. It turns out that this adjustment is important when the labor market is tight.
During periods of low unemployment, young people who stay employed tend to experience larger proportionate wage bumps than older workers. In 2017, for example, the weighted median is 40 basis points higher than the unweighted version. However, both the unweighted version (the gray line in the chart below) and the weighted version of the WGT (the blue line) declined by a similar amount from 2016 to 2017. The decline in the weighted median is also statistically significant (the p-value for the test is 0.07, indicating that the observed difference is unlikely to be due to chance).
Another issue that could affect comparisons of wage growth over time is the changing demographic characteristics of the workforce. In particular, we know that workers' wage growth tends to slow as they approach retirement age, and the fraction of older workers has increased markedly in recent years. To examine this trend, we re-computed the weighted median, but fixed the demographic and job characteristics of the workforce so they would look as they did in 1997.
Our 1997-fixed version shows that median wage growth in recent years would be a bit higher if not for the aging of the workforce (the dashed orange line in the chart below). Moreover, this demographic shift appears to explain some of the slowing in median wage growth from 2016 to 2017. Whereas the 1997-fixed median also slows over the year, the difference is not statistically significant (a test of the null hypothesis of no change in the 1997-fixed weighted median between 2016 and 2017 yielded a p-value of 0.38).
Long story short, our analysis suggests that median wage growth of the population of wage and salary earners is currently higher than the WGT would indicate, reflecting the strong wage gains young workers experience in a tight labor market. Moreover, the increasing share of older workers is acting to restrain median wage growth. Although the decline in median wage growth from 2016 to 2017 appears to be partly the result of the aging workforce, there still may be more to it than just that, and so we will continue to monitor the WGT and related measures closely in 2018 for signs of a pickup. We also want to note that with the release of the February wage data in mid-March, we will make a monthly version of the weighted WGT available.
- Does Loyalty Pay Off?
- Immigration and Hispanics' Educational Attainment
- Are Tariff Worries Cutting into Business Investment?
- Improving Labor Market Fortunes for Workers with the Least Schooling
- Part-Time Workers Are Less Likely to Get a Pay Raise
- Learning about an ML-Driven Economy
- Hitting a Cyclical High: The Wage Growth Premium from Changing Jobs
- Thoughts on a Long-Run Monetary Policy Framework, Part 4: Flexible Price-Level Targeting in the Big Picture
- Thoughts on a Long-Run Monetary Policy Framework, Part 3: An Example of Flexible Price-Level Targeting
- Thoughts on a Long-Run Monetary Policy Framework, Part 2: The Principle of Bounded Nominal Uncertainty
- August 2018
- July 2018
- June 2018
- May 2018
- April 2018
- March 2018
- February 2018
- January 2018
- November 2017
- October 2017
- Business Cycles
- Business Inflation Expectations
- Capital and Investment
- Capital Markets
- Data Releases
- Economic conditions
- Economic Growth and Development
- Exchange Rates and the Dollar
- Fed Funds Futures
- Federal Debt and Deficits
- Federal Reserve and Monetary Policy
- Financial System
- Fiscal Policy
- Health Care
- Inflation Expectations
- Interest Rates
- Labor Markets
- Latin America/South America
- Monetary Policy
- Money Markets
- Real Estate
- Saving, Capital, and Investment
- Small Business
- Social Security
- This, That, and the Other
- Trade Deficit
- Wage Growth