Skip to content

Posts by Polina Vlasenko, PhD, Senior Research Fellow

Best of 2016: What Should the Minimum Wage Be?

This week we are revisiting some of the best Daily Economy blogs of 2016. This piece first ran in May.

In recent years we’ve heard many arguments in favor of raising the federal minimum wage significantly above its current level of $7.25 an hour. Some states (New York and California are the largest example) have adopted legislation mandating an increase in the state minimum wage. In most cases, the proponents of a higher minimum wage argue that it should be set at $15 an hour. But why $15?

One argument often given for raising the minimum wage is that inflation erodes the real value of the wage over time. If one were to adjust the minimum wage for inflation, today’s wage is below what it was decades ago. Thus, the argument goes, the wage should be adjusted to restore its real purchasing power.

The adjustment for inflation is done on the chart below, for the entire period the federal minimum wage existed, using constant 2015 dollars.

Today, the real value of the minimum wage is indeed below what it was in the 1960’s. The minimum wage reached the highest real value in 1968. That year the minimum wage was raised to $1.60, which is equivalent to about $10.80 in 2015. Thus, if the objective is to match the highest-ever real value of the minimum wage, today’s minimum wage would need to be raised to $10.80 an hour. Raising it to $15 an hour would far exceed any past level of real minimum wage.

But why should we raise the minimum wage to match its past real values? In general, workers are not guaranteed that their wages will rise with inflation to preserve their real value, and there is no reason why minimum wage workers should be an exception.

Moreover, other wages are often affected by the changes in the minimum wage. Economists, business owners, and managers long knew that employees tend to pay attention to their relative pay, not only the wage itself. An experienced employee who has been with a company for a few years might feel slighted if he is paid the same wage as the newly hired inexperienced person, even if that wage is $15 an hour. That’s why some argue that increasing the minimum wage would also push up the other wages, higher up the pay scale.

If the question is whether minimum wage workers have lost ground compared to their higher-paid counterparts, let’s look at the historical trends.

The chart below compares the federal minimum wage to the average weekly earnings of production and nonsupervisory employees (the closest approximation to hourly employees that can be found in the data).

Manufacturing provides the longest history of the average hourly earnings. Other sectors have shorter data series, but the overall trend is the same.

Throughout the 1950’s and 1960’s, the average hourly earnings in manufacturing were about twice as high as the minimum wage. In mining and in construction, the average hourly earnings were about 2.5 times higher than the minimum wage.

From the late-1960’s to the early-1980’s the ratio of average hourly earnings to the minimum wage was rising. But since the mid-1980’s that ratio appears to have stabilized. In mining and construction industries that ratio is about 3.5 – the average hourly pay is about 3.5 times higher than the minimum wage. The average hourly pay in manufacturing and in the service providing industries is about 2.8 times higher than the minimum wage.

We can conclude from this chart that the growth in the minimum wage did, in fact, fall behind the growth of other wages, since in recent decades the average hourly earnings exceeded the minimum wage by more than they did in the 1950’s and 1960’s. The minimum wage workers did not see the same wage growth other workers did.

But did the minimum wage fall behind enough to justify an increase to $15 an hour?

If an objective of policy were to rectify the situation and increase the minimum wage to restore it to the same position relative to other wages it enjoyed in the 1950’s and 1960’s, how high would the minimum wage need to be? Since different industries experienced different paces of wage growth, as is evident from the chart, the answer would differ somewhat, depending on the industry we choose to look at.

But the differences are not drastic – the minimum wage in the range of $10.00-$10.60 an hour would be sufficient to take the ratio of average hourly earnings to the minimum wage back to the level seen in the 1950’s, the time when minimum wage was closest to the average wages.

The data suggest that a minimum wage somewhere between $10 and $11 an hour is sufficient to both restore its position relative to other wages in the economy, and to restore its real purchasing power back to the historical heights seen in the 1960’s. Raising the minimum wage to $15 an hour goes far beyond any past historical experience.

Click here to sign up for the Daily Economy weekly digest!

Sobering Signs in the Jobs Report

The U.S. economy added 151,000 jobs in August, according to the report released this morning by the Bureau of Labor Statistics. This represents a slowdown in the pace of job creation compared to the previous two months, when payrolls grew by over 250,000 jobs in each month. So far in 2016, job growth averaged 182,000 per month, slower than in either of the previous two years. (See chart 1).

Of the 151,000 jobs added in August, the vast majority were in private service providing industries, which added 150,000 jobs. Goods producing industries cut 24,000 jobs from their payrolls, mostly in durable-goods manufacturing, which lost 16,000 jobs. Construction lost 6,000 jobs, a fifth month of weak jobs growth, after a fairly strong period in late 2015-early 2016. The continued contraction of jobs in these sectors, which are most sensitive to changes in consumer demand, is a worrisome sign since it may indicate slowing consumer demand overall. As we have said before, the economy has been expanding, albeit slowly, on the strength of the American consumer.

The government added 25,000 jobs, almost all of them in local government, equally split between education-related jobs and non-education.

Within service industries, the largest payroll gains happened in food services and drinking places (+34,000 jobs), social assistance (+21,700 jobs), professional and technical services (+20,100 jobs), retail trade (+15,100 jobs), finance and insurance (+14,400 jobs) and health care (+14,400 jobs).

Jobs in employment services and temporary help agencies shrunk in August. This could be a good sign if it was the result of people getting into permanent positions instead of temporary jobs. However, with the subdued jobs growth overall, an alternative interpretation—that the slowdown in employment services reflects the slow labor demand overall—is at least equally likely.

The unemployment rate remained unchanged in August at 4.9 percent, a level most economists would consider to be close to full employment. The labor force participation rate, at 62.8 percent, and the employment-to-population ratio, at 59.7 percent, also remained unchanged.

A good sign is that the average duration of unemployment has shrunk: In August it was 27.6 weeks, down from 28.1 weeks in July. While the change seems small, the good thing is that it occurred because the number of the long-term unemployed has decreased considerably. The number of those unemployed 15 to 26 weeks fell by 94,000, and those unemployed for 27 weeks or longer fell by 14,000.

At the same time, the growth of employee earnings slowed down (see chart 2). In August, the average hourly earnings of all employees in the private sector were 2.4 percent higher than they were 12 months ago. The average weekly earnings, however, were only 1.5 percent above their value a year ago due a decrease in average weekly hours.

Chart 2:

Overall, despite the low unemployment rate, the latest labor market data do not appear to be overly positive. With job losses in the most demand-sensitive sectors and subdued jobs growth overall, the improvement in the labor market appears to be proceeding only slowly. Such an environment does not provide strong reasons for the Federal Reserve to raise interest rates when the Fed officials meet later this month.

Click here to sign up for the Daily Economy weekly digest!

What Gets Lost in the Income Inequality Debate

Recently, the Congressional Budget Office published an interesting study – “Trends in Family Wealth, 1989 to 2013.”

The main findings of the study can be summarized in a few points:

  • Wealth inequality has increased between 1989 and 2013 (more on this below)
  • The latest recession (2007-2009) hurt the wealth of most families, but especially those at the bottom of the wealth distribution
  • Education matters – the higher the educational attainment of the person heading the family, the higher the family wealth (no surprise here)
  • Wealth is accumulated over time – family wealth rises with the age of the person heading the family (again, not a surprise)

However, the most interesting insight from this report is that while the richest Americans are getting richer, that’s not the whole story.

To understand what this means, let’s look a little deeper at the data.

The study uses data from the Survey of Consumer Finances, a triennial survey of U.S. families sponsored by the Federal Reserve. The survey collects information about various financial matters, including peoples’ access to bank accounts and credit cards, their financial assets and non-financial assets (home equity and business equity, for example), the debt they carry, and their income.

Based on this information, the CBO constructed their measure of wealth as “marketable wealth, which consists of assets that are easily tradable and that have value even after the death of their owner.”

This excludes one important source of financial resources many families have – Social Security benefits and defined benefit pensions. (Not many families have defined benefit pensions these days, but quite a few receive Social Security benefits, either retirement or disability.) Because these things do not have value after the death of their owner, they are not included into the measure of wealth.

But they may constitute a significant share of financial resources: Just imagine what Social Security benefits (retirement or disability) mean to a low-income household that holds few, if any, other assets. This likely goes a long way to explaining how families in the bottom 25 percent of wealth distribution survive, despite the fact that their average wealth is negative throughout the period studied.

The CBO measures wealth as assets minus debt, which is the right way to go. Debt includes “nonmortgage debt, including credit card debt, auto loans, and student loans, for example.” Mortgage debt is not subtracted because on the asset side only the home equity (not the gross value of the home) was counted.

Here is the chart from this study that is likely to gain the most attention:

Most people would interpret this chart as showing the rich are getting richer. And it would appear that way. In 1989, the top 10 percent of earners owned 67 percent of the wealth. By 2013, that grew to 76 percent.

Well, it turns out there is another way to describe it, or at least a more nuanced view of it. The heart of the matter is in the phrase that CBO put in parentheses next to the chart: “Families in the top 10 percent of the wealth distribution in 2013 were not necessarily the same as those in the top 10 percent in earlier years.”

When looking at charts like these, people commonly make the assumption that “the rich” on top of the distribution are the same people throughout. But, of course, they are not.

Here is an example to drive the point home. Mark Zuckerberg is one of the very rich people today. What do you think his wealth was in 1989?  Given that he was born in 1984, there was no such thing as “his wealth” in 1989. He was part of a family that fell somewhere on the wealth distribution, but it was definitely not in the same place of the distribution that he occupies today.

Zuckerberg might be an extreme example, but it illustrates a general point – the distribution of wealth and income does not stay static over time. There is at least some amount of churn, changing who is “on top” at any given time. At least in the case of income, people have tried to describe this churn. For example, this study uses a very large data set to describe what happens to the income of top earners (top 10 percent, top 1 percent, and top 0.1 percent of income distribution over time). When I saw an early presentation of this research at a conference in 2014, the authors emphasized that the turnover at the top of the income distribution was significant from year to year.

Wealth distribution is probably a bit more stable than income distribution, but there is some turnover there as well. I remember a story in 2009 about John McAfee, the founder of the McAfee software (you probably have their anti-virus software on your computer), which reported that he had lost most of his $100-million fortune and was reportedly worth only $1 million. Now, $1 million would still put him in the top 10 percent of wealth distribution, but the point here is that even a $100-million fortune can disappear in a few years.

So, here is a side note to interpreting charts like the one above that you do not hear often. It does NOT mean that “those who were rich to begin with will remain rich forever.” Some will, but some won’t. What is does mean is “Those who ended up being rich in the end (in 2013) were a lot richer than the rich people before them.”

Maybe this is not a surprise: Those who ended up on top of the wealth distribution in 2013 had to outdo those who were on top before them.

This interpretation also illuminates one more chart from the CBO report:

One might interpret this chart as saying that the wealth of those in the 90th percentile was not hurt in the latest recession, but the wealth of everyone else was. But that is not the case (and the story of John McAfee is only one example of this). What the chart does say is: Those who somehow managed to protect their wealth from the effect of the latest recession stayed in the 90th percentile (or rose to it), but those whose wealth was damaged significantly by the recession fell to lower percentiles.

There is no hereditary entitlement that once you get to the 90th percentile of the wealth distribution, you can never fall from it. People can, and do, move down the wealth distribution. And they can, and do, move up.

Click here to sign up for the Daily Economy weekly digest!

Elections, Presidents, and the Economy

The election season will soon be heading into the final stretch. The state of the economy and economic policies, actual or proposed, are almost always major topics of discussion and debate. This election cycle is no exception. Donald Trump gave his major economic speech last Monday.  Hillary Clinton has criticized his proposed policies.

But how much can a president really affect the country’s economic performance? The forces that affect the economy act over long periods of time, and in many cases there is not much that a president can do, no matter his or her economic agenda.

Robert Gordon, a prominent economist whose research focuses on issues of economic growth, suggests as much. In a recent op-ed he argues that the slower growth and wage stagnation that make so many voters worried about their economic prospects are “trends that are decades in the making,” and they may not be easily changed by presidential policies.

For example, in the past four years, GDP has averaged only 2 percent per year, much below the over-3-percent growth seen for decades prior to 2007. But that is constrained by deep economic undercurrents. The labor force is not expanding as quickly as it did in decades past because, among other things, female labor force participation has likely reached its peak, and has levelled off. The aging of the population and the anticipated retirement of the Baby Boomers is slowing labor force growth as well. It is unlikely that any president or any policy can alter these trends.

With the labor force growing slowly, the only way to increase output is to somehow generate faster productivity growth. Productivity growth is also crucial for wages, because wages can grow only if output per hour is rising. But here, too, the U.S. economy is facing difficulties. The latest productivity numbers show that productivity (output per hour in the nonfarm business sector) declined in the second quarter of 2016. In recent years, productivity has grown slowly, averaging just one percent annual growth since 2010. This contrasts with much faster productivity growth in earlier decades (see chart).

aug 16 chart 1 650 pix

In his op-ed, Gordon discusses several reasons for the slowdown in productivity. He is known in the economic profession for putting forward a view that most of the “low hanging fruit” of productivity enhancing inventions have already been picked. Gordon says we are in for a long period of much slower productivity growth. If this is the case, policies, or presidents, have limited ability to affect such trends.

Nevertheless, presidents will forever continue to claim credit for the economic successes that happen during their tenure, and their opponents will point out economic calamities that occur on their watch. And the presidential candidates will always claim that they have the recipe for improving the economy. This happens because the voters would like to hear it, not because it is, or can be, supported by data.

Here is a look at the economic performance under past presidential administrations.

aug 16 table 1 650 pix

Many comparisons can be made, but overall it is difficult to argue that one party systematically performs better than the other as far as the economy is concerned.

Click here to sign up for the Daily Economy weekly digest!

Job Growth Improves

The Employment Situation Report released by the Bureau of Labor Statistics this morning shows improvement in the labor market compared to the gloomy report last month. According to the latest data, in June the U.S. economy added 287,000 jobs, but the May numbers were revised down to only 11,000 jobs from the 38,000 reported a month ago.

This brings the average growth so far in 2016 to 172,000 jobs per month. Despite the improvement in June, the average is still below the annual averages seen in the past two years. (See chart).

chart 1 july 650 pix

Almost all job growth came from private service-providing industries, which added 256,000 jobs. Within the service industries, the largest increases came in leisure and hospitality (59,000 jobs added), health care and social assistance (58,400 jobs), and information (44,000 jobs, but most of this reflects the return of workers from a strike). Goods-producing industries added only 9,000 jobs and government added 22,000.

At the same time, the unemployment rate increased in June to 4.9 percent, from 4.7 in May. This, however, is not a bad sign, because the increase in the unemployment rate came primarily from more people entering the labor force. The civilian labor force grew by 414,000 in June, and the labor force participation rate rose to 62.7 percent from 62.6 percent. For both measures, this is the first increase since March of this year.

Most of the people who entered the labor force, however, remained unemployed in June; employment rose by only 67,000. Still, people flooding back into the labor force is usually a sign that they see improved prospects of finding jobs. The fact that, in June, many people apparently judged their job prospects as improved is a positive development for the economy overall.

Improved jobs numbers in June suggest that May’s dismal report may have been a one-time anomaly. Time will tell. For now, it appears that the labor market remains a point of relative strength in the economy.

Click here to sign up for the Daily Economy weekly digest!

New GDP Data Reveal Slightly Faster Growth

This morning’s release from the Bureau of Economic Analysis reveals that U.S. economic growth in the first quarter of 2016 was a bit faster than we previously thought. U.S. gross domestic product, the broadest measure of the economy’s output, grew at 1.1 percent (annual rate) in the first three months of the year, somewhat faster than the 0.8 percent growth reported in the previous estimate a month ago. But still, the pace of growth is considerably slower than the average growth in the past two years (see Chart 1), leaving unchanged the overall picture of a slowdown in growth at the end of last year and the start of this one.

chart1_june 650p

Today’s estimate is based on more complete data than the earlier ones, but it is still not the final word on GDP growth. Next month the BEA will release its annual revision to the National Income and Product Accounts that can potentially affect the estimates of GDP growth for the past three years.

Two main factors are behind the latest estimate showing faster GDP growth than was previously believed: business investment did not shrink as much as we originally thought, and exports did considerably better than previously reported. The newest estimate shows that, in the first quarter of 2016, personal consumption expenditures grew somewhat more slowly than was previously believed; investment shrank, but not as much as we originally thought; exports increased instead of falling; imports (which are subtracted from GDP) fell more than we previously thought; and government spending increased about as much as in the earlier estimates. See Chart 2:

chart2_june 650p

These revisions were necessary because earlier estimates had to rely on incomplete data, making assumptions and extrapolations for the missing information. Once more complete data are available, the previously extrapolated values are replaced by true ones. Thus, the revisions tend to be more substantial in those areas where the data take a while to arrive, and where it’s more difficult to fill in the missing information.

Exports is one category where extrapolating missing data accurately is difficult. This month, exports posted the largest revision of all GDP components: The latest data show that exports rose 0.3 percent in the first quarter of 2016, while the earlier (incomplete) data indicated that exports fell 2 percent.

Another place where extrapolations of missing data is often difficult is within the investment category. The latest data show that private domestic investment (which includes business investment and residential construction) fell 1.8 percent (annual rate) in the first quarter of 2016. This is a smaller decline than the 2.6 percent drop reported earlier. The difference between these two estimates comes almost entirely from business investment, and specifically investment in intellectual property products. It turns out that businesses have increased their spending on software and research and development much more than was previously estimated.

Overall, while this latest estimate paints a picture that is less gloomy, the pace of economic growth early in the year remains quite slow, raising concerns for the vitality of the expansion.

Click here to sign up for the Daily Economy weekly digest!

A Worrying Sign on Jobs

In a year when economic growth has seemed tenuous, the resilient labor market had been providing welcome reassurance.

But the report released by the Bureau of Labor Statistics this morning shows that the U.S. economy added only 38,000 jobs in May, far below expectations. The headline number was reduced by 35,100 due to striking workers at Verizon Communications. Without the strike, payrolls would have increased by 73,000. Either way, the jobs growth in May is much slower than it was earlier in the year (See Chart 1).

Chart 1:

Yet another worrying sign is that the weakness in job creation appears to be spreading across industries (see chart 2).

Chart 2:

Earlier this year, only mining and logging, durable goods manufacturing, and transportation were posting job losses. This was easily explained by shrinking domestic oil production as a result of a sharp decline in crude oil prices. Employment in oil extraction (part of the mining and logging sector) was shrinking, as was production of heavy equipment used in the oil sector, leading to job losses in durable goods manufacturing. Transportation of domestic crude oil (much of which is done by rail) was falling as well.  There was a possibility that the weakness and the job losses would be contained in the industries affected by crude oil production.

But in April and May, it seems that the job losses spread to other industries. Now, in addition to mining, durable goods manufacturing and transportation, job losses have appeared in construction, wholesale trade, utilities, retail sales, and information services (although the last one is mostly due to a strike). Other industries continue to add jobs, but at a much slower pace than before. This is a worrying sign.

And yet, the unemployment rate still fell in May to 4.7 percent, from 5 percent in April. How can this be? The answer is not encouraging – the drop in the unemployment rate is mostly driven by people dropping out of the labor force. In May, 458,000 people left the labor force, after 362,000 did so in April. Such a substantial decrease in the labor force for two months is a row is unusual (see chart 3) and is yet another worrying sign.

Chart 3:

To sum up, this was a very disappointing employment report with several worrying signs for the economy. It makes it nearly certain that the Federal Reserve will not raise interest rates when it meets in June. We will be watching other data carefully to see if the worrying signs seen here are confirmed by other information.

Click here to sign up for the Daily Economy weekly digest!

A Winter of Less Discontent Than We Originally Thought

New data out this morning show it appears that the economy’s slowdown earlier this year was not as drastic as we initially concluded.

This morning, the Bureau of Economic Analysis released the revised estimates of the gross domestic product, the broadest measure of economy’s output, for the first quarter of 2016.  According to the revised data, the GDP rose 0.8 percent (annual rate) in the first quarter of 2016, somewhat faster than the 0.5 percent rate indicated by the initial estimate a month ago. In the final quarter of 2015, GDP grew 1.4 percent. These estimates may yet change again when the final estimates of GDP are released next month.

The revisions were modest:

But the details provide some reassurance about the state of the economy, which has given us reason for concern in recent months.

Economic growth remains primarily driven by consumer spending, which rose 1.9 percent (annual rate) in the first quarter. Consumer spending is likely to remain strong in the near future. Today’s data release shows that disposable personal income posted a 4 percent increase in the first quarter of 2016, up from the 2.9 percent increase reported earlier. This is a faster pace of growth than at any time in 2015. A healthy growth in disposable income, which represents the fund people have available after all taxes have been paid and transfers, if any, received, makes the continued growth in consumption expenditures possible, suggesting that economic growth will continue.

Gross private domestic investment – which includes business investment and residential construction– performed a bit better than the original estimate indicated. The revised estimate still shows a decline of 2.6 percent (annual rate), but this decline is milder than the 3.5 percent drop reported earlier. The improvement came primarily from the faster growth in residential investment (housing), which rose 17.1 percent, faster than the 14.8 percent increase reported earlier. Business investment, on the other hand has shrunk 6.2 percent in the first quarter. If business investment returns to growth in the future, we should see significant improvements in the overall GDP numbers.

Net exports also showed a small improvement compared to the initial estimate. Exports shrunk a bit less: In the first quarter they fell 2 percent instead of the previously reported 2.6 percent. Imports, which are subtracted from GDP, fell 0.2 percent according to the new data, instead of the previously reported increase of 0.2 percent. As a result, net exports subtracted only 0.2 percentage points from GDP growth, while in the earlier report it subtracted 0.3 percent. Net exports are restrained by the strengthening U.S. dollar, which makes U.S.-made products more expensive for foreign buyers. The strength of the dollar is likely to continue in the near future, which suggests that the growth in net exports will remain subdued.

Click here to sign up for the Daily Economy weekly digest!

What Should the Minimum Wage Be?

In recent years we’ve heard many arguments in favor of raising the federal minimum wage significantly above its current level of $7.25 an hour. Some states (New York and California are the largest example) have adopted legislation mandating an increase in the state minimum wage. In most cases, the proponents of a higher minimum wage argue that it should be set at $15 an hour. But why $15?

One argument often given for raising the minimum wage is that inflation erodes the real value of the wage over time. If one were to adjust the minimum wage for inflation, today’s wage is below what it was decades ago. Thus, the argument goes, the wage should be adjusted to restore its real purchasing power.

The adjustment for inflation is done on the chart below, for the entire period the federal minimum wage existed, using constant 2015 dollars.

Today, the real value of the minimum wage is indeed below what it was in the 1960’s. The minimum wage reached the highest real value in 1968. That year the minimum wage was raised to $1.60, which is equivalent to about $10.80 in 2015. Thus, if the objective is to match the highest-ever real value of the minimum wage, today’s minimum wage would need to be raised to $10.80 an hour. Raising it to $15 an hour would far exceed any past level of real minimum wage.

But why should we raise the minimum wage to match its past real values? In general, workers are not guaranteed that their wages will rise with inflation to preserve their real Read more

Economic Weakness Appears to be Spreading

Yesterday’s report from the Bureau of Economic Analysis gives the first look at the aggregate state of the economy this year, and it provides reasons for concern.

According to the report, in the first three months of 2016, real gross domestic product, the broadest measure of the economy’s output, grew at a 0.5 percent annual rate. This is the slowest pace of growth in two years (see chart 1).

Yesterday’s report presents what is known as an advance estimate of GDP, which is based on data that are incomplete and subject to further revisions in the coming two months. The revisions could change the picture considerably.

In the past, the size of revisions from the advance to the final (third) estimate of GDP averaged 0.6 percentage points. This suggests that, by the time we have the complete data, the current estimate of 0.5 percent GDP growth in the first quarter of 2016 can easily change to anywhere between -0.1 percent growth (i.e. a reduction in GDP) to 1.1 percent growth.

Whether the future revisions change the overall picture, there are reasons for concern within yesterday’s report. A closer look at the numbers shows that most of the cyclically-sensitive components of aggregate demand decreased in the first quarter of 2016. That is, the types of spending that react to the worsening of economic conditions early and most strongly are falling. These include: consumer spending on durable goods (down 1.6 percent annual rate in the first quarter), business fixed investment in equipment (down 8.6 percent) and structures (down 10.7 percent).

The accumulation of inventories by businesses has slowed to the lowest pace in two years. And even within consumer spending on nondurable goods (a category that is not very sensitive to cyclical factors), spending on things that are discretionary, such as clothing and footwear, has gone down.

All these categories of spending contributed negatively to GDP growth in the first quarter. Another sizeable negative contribution came from exports, which shrunk 2.6 percent (annual rate), likely due to the stronger U.S. dollar and somewhat weaker global growth. A decrease in almost all cyclically sensitive components of spending suggests that the weakness in the economy is spreading.

Still, the overall GDP grew in the first quarter of this year. Two things are fueling this growth: healthy growth in consumer spending on services (up 2.7 percent annual rate), and an increase in residential investment, or housing construction (up 14.8 percent).

A robust growth in housing, usually a cyclically sensitive component, in the face of a decrease in all other cyclically sensitive spending, may look surprising. But it really isn’t.

The Fed’s policy of exceptionally low interest rates has substantial influence on the housing market. In an extremely low-interest-rate environment, residential investment can continue to grow even when the economy enters a recession. This happened in 2001, when U.S. economy went through a mild recession, but the housing sector never knew it, fueled by the then-unprecedented extra-low interest rates maintained by the Fed. Today, the Fed maintains interest rates much below what we’ve seen in 2001. Thus, rising residential investment and housing construction is no surprise, but it might not reflect the strength of the consumer as much as it reflects the low interest rates.

There is one indication of the strength of consumers in yesterday’s report. The real disposable personal income, the money people have available after all taxes are paid and any transfers received, continues to grow at a decent rate. In the first quarter of 2016 it grew at a 2.9 percent annual rate, and was 2.8 percent higher than a year ago, a faster pace of growth than that of GDP.

This means that consumers do have income to spend. They just seem to be reluctant to do so – the saving rate ticked up to 5.2 percent of disposable income, from 5 percent in the last quarter of 2015. In general, the saving rate has been elevated in this business-cycle expansion compared to the previous one (see chart 2), which might explain the subdued growth in GDP. A higher saving rate is good for each individual’s financial security, but if most consumers decide to save more, it translates into restrained spending for the overall economy.

The revised GDP estimates that will come out in the next two months will give us a better idea of whether this growth slowdown is temporary or not. One thing is certain, however – the growth of U.S. economy seems to be set on a slower trend for a while. When the recession ended in 2009, the economy’s growth never returned to the pace seen in other expansions. Since 2009, the U.S. economy grew, on average, around 2 percent per year (see chart 1 above). This is far below the average growth in past business-cycle expansions, which reached around 4 percent. The current slowdown may be temporary, but the unimpressive trend growth of about 2 percent per year seems to be here to stay.

Click here to sign up for the Daily Economy weekly digest!