Skip to content

Posts from the ‘Jobs’ Category

Teaching High School Students About Unemployment


How would you use a sampling distribution to measure unemployment? Students at Lee Middle and High School, in Lee, Massachusetts, recently learned this lesson with some help from the American Institute for Economic Research.

Teachers who participated in AIER’s Teach-the-Teachers Initiative are field testing their lesson ideas. On January 28, we visited Lee to observe the AP Statistics class taught by Thomas McCormack. Prior to our visit, Mr. McCormack assigned each of his 20 students a labor force status (for instance, a retiree, a full-time student, a laid-off person). In addition, each student was given a randomly generated list of 10 of their classmates. This was an attempt to replicate the Bureau of Labor Statistics’ monthly survey when they use random digit dialing. Thus, each student had a sample of 10 students and was told to record the labor force status of each student in their sample. Then each student calculated the unemployment rate for their sample by dividing the number of unemployed by the number of people in the labor force, which is the sum of employed and unemployed. So each student obtained one number to represent the unemployment rate for their sample.

During the class period each student came up to the white board and put their unemployment rate statistic on a number line. As a whole, the class created a dot plot of 20 unemployment rates. Of course, each of the 20 unemployment rates was different, and represented a distribution of unemployment rates in 20 possible samples. This is exactly the definition of the sampling distribution! It was visual, it was understandable to the students, and it was cool!

Mr. McCormack then posed the question: “We have only one true number of the unemployment rate in this population, but we have all these dots here…. Why do you think that is?” The students discussed the concepts of a “range,” a “mean,” and a “standard deviation.” The teacher went on to describe the normal distribution and its properties.

IMG_1021 Based on the sampling distribution of the unemployment rate depicted on the board (in this class we had rates that were as low as 17% and as high as 57%), students predicted the unemployment rate for the whole population and then compared this prediction with the actual unemployment rate defined by the teacher for this activity. This helped them to see that, even though there was one true unemployment rate for the population (33%), individual samples can vary quite a bit. The class discussed the reasons for the “errors” and proposed ways to minimize them, such as using a larger sample, for example.

This creative lesson required students to actively engage with the concepts of an unemployment rate and a sampling distribution. In addition to employing the inquiry-based method, it allowed the infusion of economic concepts into a statistics class, promoting an interdisciplinary approach to teaching and learning. All of these things are important for building college- and career-readiness skills among high school students. We are happy that our program helped Mr. McCormack to create such a stimulating lesson.

We are offering three workshops for teachers during the Summer 2016. Visit our Web site to learn more and to register for the program:

Pictures: First picture shows the role-playing cards being used by students to guess the actual unemployment rate after they saw the sampling distribution on the board, which is shown on the second picture.


AIER Issue Brief: H-1B Visas Have no Impact on Average Wages

How H-1B visa holders impact American wages is hotly debated. Many argue that H-1B workers, the foreign-born, highly skilled, temporary workers that companies can hire for three to six years, drive down American wages. Others argue that H-1B workers actually help increase wages for Americans.

But what if H-1B visa status has less impact on wages than we think?

Read more

Marketplace of Ideas


Money Talks Illustration by Oliver Munday, The New Yorker

GDP grew more than expected last quarter, jobless claims remain low, and one survey-based index says manufacturing activity hit a three-year high in July. Here are a few other stories from the week’s economic news:

  • The verdict on the July jobs report is that it was a “mild disappointment.” Non-farm payroll employment rose by 209,000 last month, below gains of 298,000 in June and 229,000 in May. Since the start of the year, the labor force has grown by an average of 126,000 people per month. Over the same period, job gains have averaged 230,000 per month. If sustained, that difference is enough to keep the unemployment rate falling by more than 0.1 point per month. The unemployment rate did nudge up a bit in July, to 6.2 percent from 6.1 percent, but that follows an outsized 0.6-point drop between March to June. While unemployment is down from a peak of 10.0 percent during the recession, Justin Wolfers, writing for the Upshot, says it is still way too high. Wolfers observes that if you were to line up all the unemployed people in the country, they would stretch from New York to San Francisco. That is a striking visual, but what does it mean for the economy? Following the recession, unemployment peaked at 15 million people. That number is down to 9.7 million now, but it is still higher than the low of 6.8 million prior to the recession. AIER’s Business-Cycle Conditions analysis suggests the economy will strengthen in the second half of the year, which should keep the employment picture improving.

Read more

Why Allowing H1B Spouses to Work May Only Increase the Tech Gender Divide

Ironically, allowing H1B spouses to work in the U.S. might only exacerbate the gender divide in the U.S. computer technology industry. The White House recently released a statement that proposed giving work visas to spouses of H1B visa recipients.

As it stands now, the H1B visa program, or the program that allows high-skilled temporary workers entrance to the U.S., does not provide a work visa for recipients’ spouses. Spouses currently can get a student visa, but unless they are on their own work visa, they legally cannot work.

We have heard that this will make the already highly demanded visa even more desirable, especially for computer technology workers. AIER’s latest Research Brief, H1B Jobs: Filling the Skill Gap, showed that computer technology requests make up the lion’s share (70%) of H1B petitions.

Marital Status

Source: AIER Analysis of Ruggles, Alexander, Genadek, Goeken, Schroeder, and Sobek. Integrated Public Use Microdata Series: Version 5.0 [Machine-readable database]. Minneapolis: University of Minnesota, 2010.

What will this mean for H1B visa holders? AIER’s analysis from 2010-2012 Census data shows that in the computer technology industry, there are larger shares of H1B men that report being unmarried compared to U.S. citizen men: 74% of H1B men are unmarried, compared to only 6% of U.S. citizen men.

But many H1B females in computer technology are married. 76% of H1B females are married, compared to 58% of U.S. citizen females in that field.

Right now, H1B women have the highest rates of educational attainment in the U.S. computer technology industry (94% have a college degree or higher), but they make up the smallest share of H1B holders in the industry (22%).

It is safe to assume that allowing spouses of H1B visa holders to work will increase the numbers of married applicants. But since the majority of H1B women are already married, this policy might only further exacerbate the gender disparity among H1B visa holders as more men than women apply.


Three Takeaways from the April Employment Report

The monthly Employment Situation Report for April from the Bureau of Labor Statistics was followed up by a deluge of press coverage. How can you make sense of the data itself, let alone the cacophony in the press? We have three key takeaways for you:

1) A tightening labor market doesn’t seem to be sparking inflation.

The overall unemployment rate plunged to 6.3 percent in April from 6.7 percent in March. Economists agree that the unemployment rate affects inflation, but they disagree about the nature of that relationship. The Fed estimates the lowest the unemployment rate can go without causing accelerating inflation is about 5.5 percent. But a recent flurry of research suggests it’s the short-term unemployment rate, covering workers unemployed for fewer than six months, that affects inflation, because those out of a job for longer than six months are on the “margins” of the labor market.

That short-term rate fell to 4.1 percent in April from 4.3 percent in March. It has been falling steadily since the end of the recession, and it’s quickly closing in on the pre-recession low of 3.7 percent. Does that mean we should be bracing for a surge in wages, with inflation hot on its heels? Perhaps not: Earlier this week, the employment cost index was reported up only 1.8 percent compared to a year ago in the first quarter, slowing from a 2.0 percent pace in the fourth quarter. For the last four years, the index has been pretty rock solid, wobbling in a tight range between 1.7 and 2.0 percent. No upward pressure there. Meanwhile, the core PCE price index—one of the Fed’s key inflation measures—rose just 1.2 percent compared to a year ago in March. That’s up slightly from 1.1 percent in January and February but far from the 2.4 percent peak pace seen before the recession.

The short-term unemployment rate may have been a better guide to inflation pressures than the overall unemployment rate in the past, but a seemingly rock-bottom short-term rate now isn’t lighting any fires under wages or prices.

2) There may not be any hidden slack in the labor market.

Perhaps we’re not seeing bigger wage gains as unemployment falls because there’s still a lot of “slack” in the labor market. That has been Fed Chair Yellen’s view—you can see our updated “Dashboard” of her key labor market indicators here.

One of Yellen’s theories is that the deep recession caused many unemployed workers to drop out of the labor force altogether. If that’s correct, we should see the labor force participation rate rebound as the economy recovers. But that may be wishful thinking. The participation rate has been declining steadily since the end of the recession, from 65.2 percent in January 2010 to just 62.8 percent in April. While it seemed to have stabilized earlier this year, it ticked down again in April by nearly half a percentage point. That’s the main reason for the big drop in the unemployment rate in April.

As we have been discussing in our Business Cycle Conditions report, the economy is on a slow-but-steady recovery path. Perhaps it’s unwise to expect any big swings in labor market trends this far into the recovery. 

3) Noisy data affect perceptions, but so do revisions.

Along with new data showing a smoking 288,000 gain in non-farm payrolls in April, today’s report included a total of 36,000 in upward revisions to the March and April numbers, to 203,000 and 222,000 respectively. Revisions so far this year have netted a 30,000 gain compared to initially reported results. Last year, net revisions resulted in a 31,000 gain, on average.

The New York Times yesterday ran a compelling bit of analysis that illustrates why people shouldn’t get worked up over month-to-month changes in payrolls, which are mostly just “noise.” But you should also keep in mind that the payrolls numbers we see today will definitely not be the numbers we’re looking at next month, or a year from now.

Our animated chart below shows the initially reported payrolls data over the last year versus the latest reported numbers. Animation may take a few seconds to load.

Nonfarm Payrolls

People and Businesses in Recessions

The stubbornly high unemployment following the last recession, coupled with corporate profits reaching record levels, can create a picture in the minds of people that businesses are having it easy at the expense of workers. This picture may be misleading. Recessions are also tough on businesses and business owners, but the data that reflect this are not nearly as prominent as the unemployment statistics.

The way we measure unemployment is fundamentally different from the way we measure businesses. Businesses can appear and disappear in response to economic conditions, but people tend to stick around no matter what.

There is no equivalent of a long-term unemployed worker in data on businesses. People can remain unemployed for months or years, and we will have data about these numbers. Businesses usually cannot remain without customer orders for many months and survive. Such businesses close their doors and disappear from data – data about sales, data about business profits, and other data.

Read more

Marketplace of Ideas

Take a break from the NCAA men’s and women’s basketball tournaments with these stories from the week’s economic news:

  •  The latest tome from Michael Lewis, best-selling author of The Blind Side and Moneyball, hit bookshelves, Nooks, and Kindles this week. In Flash Boys, Lewis delves into the world of high-frequency trading and the technological advances that make it possible for certain players in the stock market to “scalp” others by anticipating their orders for stocks, buying up the shares first, then selling them back at a slightly higher price. In a 60-Minutes interview on Sunday, Lewis suggested that such practices amount to “rigging” the market, and “ordinary investors” are “getting screwed.” That might conjure up mental images of a dentist in Des Moines or a plumber in Palm Beach getting his or her pocket picked by an army of sinister Wall Street Gordon Gekkos. But consider this example of a typical “scam” from page 52 of Lewis’s book: A Wall Street firm is looking to buy 10 million shares (yes, 10 million) of Citigroup stock. If high-frequency traders can see the bid coming and buy up the shares first, they stand to gain $29,000. That sounds like a lot, until you realize that the total price tag on the deal in question is about $290 million. Some are arguing that the practices the book documents are nothing new, and perhaps not even illegal—there have always been intermediaries in the market. But regardless of whether you think it’s a problem, and no matter how folksy and engaging Lewis can be in his storytelling, this isn’t a Main Street versus Wall Street issue: The “victims” losing big money are investors doing multiple hundred-million-dollar deals each day—not “ordinary” folks.
  •  Fed Chair Janet Yellen shook up markets earlier this week with what one analyst described as “one of the most dovish speeches” from a Federal Reserve official, ever. In fact, Yellen didn’t deviate much from her testimony before Congress back in February, or from the statements she offered in her press conference following the last FOMC meeting. She continues to focus on the slow progress toward recovery of certain sectors of the labor market, particularly the level of long-term unemployment—those looking for a job for 27 weeks or longer—and involuntary part-time employment—those who would prefer to work full-time but can only find part-time work. The latest Employment Situation report, published by the BLS today, shows little additional progress in those areas. The standard unemployment rate was steady at 6.7% last month, and the 192,000 rise in non-farm payroll employment was close to last year’s average monthly gain of 194,000. For an overview of all the labor market indicators Chair Yellen is watching, check out our updated Labor Market Dashboard.
  •  In his New York Times column this week, economist Paul Krugman addresses one common explanation for the stubbornly high level of long-term unemployment: the so-called “skills gap.” Krugman writes in response to an article in Politico co-authored by Jamie Dimon, embattled CEO of J.P. Morgan Chase. Dimon’s firm is promoting a $250 million initiative called New Skills at Work, meant to “give aspiring workers the training they need to meet employer demand in their community.” Krugman launches at what he sees as a blame-the-workers approach: people are unemployed not because the economy is still weak, but because they have failed to acquire the skills necessary to succeed. Krugman calls the skills gap a “zombie idea”—one that has widespread acceptance but is not supported by evidence. He argues that if there was significant unmet demand for skilled employees, we would see more upward pressure on wages. “It’s very hard to find groups of workers getting big wage increases,” he says, “and the cases you can find don’t fit the conventional wisdom at all.” It’s probably worth noting that when it comes to education and training, U.S. workers could use some help to remain competitive in the global market place: A recent OECD study ranking 34 countries worldwide shows American students are below average in math skills and no better than average in science and reading.

[Photo: theogeo/Flickr]

Good but not Great: Our View of the March Employment Report

The March Employment Situation showed improvement in hiring and hours worked, reflecting a rebound from weather-distorted data in January and February. There were also upward revisions to the prior month’s jobs numbers, adding an additional 37,000 to payrolls. For the private sector, payrolls rose 192,000 in March, bringing the three-month average to 182,000, just slightly below the 12-month average of 189,000.

On the negative side, hourly wages fell 0.1 percent for the month, putting the year-over-year gains at just 2.2 percent. Faster wage gains are the key to improving consumer confidence and accelerating consumer spending. The aggregate payrolls index, which combines employment, hours, and wages, rose 1.0 percent in March and is up 4.0 percent for the quarter – in line with growth over the past few years. The rise in the payrolls index points to continued moderate growth in personal income, which in turn should support moderate gains in consumer spending.

In total, the report suggests that the first quarter ended with improving, though not great, momentum, and this positive momentum will likely carryover into second quarter GDP growth. These results are consistent with the results of AIER’s on-going business indicators analysis.

However, many of the gauges on the Yellen dashboard – particularly broader measures of unemployment and the number of marginally attached, discouraged, and involuntary part-time employees – remain weak. With the outlook for growth still considered to be below trend, and with inflation measures running below the Fed’s target, the Fed is unlikely to change the course of monetary policy based on this report.


What’s on Janet Yellen’s Dashboard?

The Federal Reserve’s Federal Open Market Committee (FOMC) this week changed its “forward guidance” on monetary policy, dropping its 6.5 percent unemployment rate “threshold” for considering an increase in the federal funds target rate. Previously, the Fed had said that it would keep its key interest rate target near zero “at least as long as” the unemployment rate remained above 6.5 percent, then “well past the time” it declined below 6.5 percent. The unemployment rate is currently 6.7 percent.

Explaining the move, Chair Janet Yellen said that as the economy comes closer to achieving full employment, the FOMC will have a more finely balanced decision about when and by how much to raise the target for short-term interest rates. So what will the Fed, and Ms. Yellen, be looking at to assess labor market conditions? In her press conference, Ms. Yellen listed a “dashboard” of indicators that she watches to gauge the progress the labor market has made in its recovery from the Great Recession. We have represented those indicators in a series of six charts, so you can get a sense of the trends and lingering troubles Ms. Yellen has her eye on.

Read more

Quitters May Indicate Rising Confidence in the Labor Market

Each month the Employment Situation from the Bureau of Labor Statistics reports, among other things, the number of jobs created and the unemployment rate. While the Employment Situation is important, the new Federal Reserve Chair Janet Yellen recently indicated that Federal Reserve officials will also focus on the lesser known Job Openings and Labor Turnover Survey (JOLTS). JOLTS reports the number of job openings, hires, and separations in the preceding month.

Job openings include full-time, part-time, permanent, and temporary positions that firms are actively recruiting for outside the company. Hires are the total number of positions filled. Separations are discharges initiated by the employer, and quits are initiated by the employee. Quits do not include retirements, transfers, or promotions.

Though there is no information on whether a quitter has a job lined up, the conventional wisdom is that workers quit when they have found, or are confident in finding, a job with higher compensation or better conditions. In March 2013, Yellen spoke at the National Association for Business Economics, stating, “A pick-up in the quit rate would signal that workers perceive their chances to be rehired as good—in other words, that labor demand has strengthened.”

The old saying “nobody likes a quitter” isn’t always true. The number of quits tend to increase as the economy expands and quitters create a dynamic and flexible economy by better matching workers to employers.

In January, there were 4.0 million job openings, 4.5 million hires, and 4.5 million separations in the U.S., of which roughly half were quits. The number of quits bottomed in August 2009, two months after the Great Recession ended. Since that time, quits have increased by nearly 50 percent as workers grew more confident in the labor market.

The quit rate, which is the proportion of quits relative to total employment, has risen from a low of 1.3 percent in April 2009 to 1.7 percent in January 2014. However, that rate is still below the average pre-recession rate of 2.0 percent, suggesting the labor market has yet to return to full strength. As Fed officials address the difficult task of tapering their asset purchases, the quit rate may be one indicator to keep an eye on.

031114 JOLTS