WHY ARE PEOPLE PESSIMISTIC ABOUT THE ECONOMY: A CLOSER LOOK AT THE DATA

 

Readers may wonder why so many have relatively unfavorable views on the US economy.  What is the source of all of this pessimism?  After all, readers may think about the positive sounding US employment report from January.  More recently, the US advance retail sales report sounded very encouraging.  Is there more to the story?  In this blog post, I offer some possible answers.

Rick Santelli of CNBC and others announced this (February 15, 2023) morning that the US Census Bureau reported a strong rebound in retail sales in the United States for January 2023. The estimated increase of 3 per cent exceeded expectations of a gain of about two per cent.  Four of the last six months had estimated decreases in retail sales, including November and December, so the January gain may have been especially welcome.  Taking automobile sales out of the analysis, retail sales were much better than anticipated, as was the case when removing automobile and gasoline sales from the analysis.  Retail sales may have been better than forecasts in all categories. 

According to an online article by Jeff Cox (2023) of CNBC, no area of retail sales had a decrease, further underscoring what may appear to be broad based strength.  Moreover, the estimated three per cent rise in retail sales between December 2022 and January 2023 outweighed the 0.5 per cent increase in the consumer price index (CPI) over the same period.  This implies an increase in not only retail sales spending, but probably also an increase in actual goods purchased.

But, some potentially critical information has not been covered yet in this blog post.  Cox (2023) points out that compared with US retail sales from January 2022, retail sales from January 2023 increased by 6.4 per cent.  Cox (2023) reports that the US CPI on an annual basis also increased by 6.4 per cent.  Thus, in terms of retail sales measured in US dollars with constant buying power, US retail sales in January 2023 were approximately unchanged from one year ago! 

Could the effects of (a) inflation, although perhaps either at a slower rate and/or a recent switch to deflation, and (b) higher interest rates be slowing consumption spending in the economy?  Jessica Dickler’s (2023) online article seems to draw that conclusion.  Among other things, Dickler (2023) notes that (1) credit card debt hit a record-setting $930.6 billion in December 2022, (2) more consumers fell behind on payments, and (3) more credit card borrowers are borrowing for shelter, food, and other necessities.  Clearly, millions of households are having difficulty keeping up with higher prices. 

Additionally, the portion of the US advance retail sales report from the Census Bureau that has generated more attention is the seasonally adjusted portion.  In fact, most if not all of the data above in this blog post are in seasonally adjusted terms.  Many economic activities decrease in January and the first quarter of a new year compared with December and the fourth quarter of the previous year.  At least in recent years, March usually has a strong rebound in retail sales with unadjusted data as temperatures warm.  Thus, it may well be appropriate to perform seasonal adjustments because a decrease in economic activity early in the new year may not mean that the economy is heading into an unusual downturn.  But with millions in the US already struggling, it may be especially helpful to look at the data on a not seasonally adjusted basis or on an unadjusted basis.

Shortly after 1:00 pm EST, Rick Santelli noted that retail sales not seasonally adjusted were much weaker than the seasonally adjusted data.  A look at unadjusted data shows that month-to-month US retail sales in January 2023 fell rather sharply last month compared with December 2022.  In fact, without seasonal adjusting, US retail sales decreased by about 17.8 per cent in January 2023 compared with December 2022.  To be sure, retail sales usually if not always decrease in January compared with the previous December.  How usual or unusual is this January decrease in retail sales in 2023 compared with other recent years?  Using data from the U.S. Census Bureau (2023) but accessed from the FRED website of the St. Louis Federal Reserve Bank, the previous estimated percentage decreases in US retail sales over the past five Januarys in reverse chronological order (from January 2022 through January 2018) are 18.1%, 17.2%, 19.8%, 19.1%, and 21.5%.   Back in 1993, 1994, and 1995, the January decreases not seasonally adjusted all exceeded 28.3 per cent.  So, one may conclude that the January US retail sales level was bad, just perhaps not as bad as one would normally expect.  In the vernacular, someone might say that the unadjusted US retail sales from January 2023 ‘sucked, but it didn’t suck as much as January normally sucks.’  If my understanding is correct that estimated 0.5 per cent increase in the January CPI was not removed from the US retail sales report to measure sales in constant purchasing power dollars, then the January 2023 retail sales report may be even more disappointing.

This is especially true when looking at the unadjusted US retail sales data for November and December of recent years.  Readers may want to note that, unlike the seasonally adjusted US retail sales data which showed estimated decreases in November and December 2022, the unadjusted data show increases of roughly 2.1 per cent and 7.8 per cent respectively.  However, in November 2021 and December 2021, the unadjusted estimated increases were about 3.9 per cent and 9.2 per cent, respectively.  In the year 2020, although November unadjusted retail sales fell by about 0.5 per cent, the estimated December 2020 retail sales increase was 13.6 per cent.  In the year 2019, the estimates for unadjusted US retail sales in November 2019 and December 2019 were 2.7 and 11.1 per cent.  For the last two months of 2018, the retail sales estimates were 4.0 and 7.4 per cent based on data not seasonally adjusted. The year before, the November and December US retail sales estimates were 5.8 and 12.4 per cent.  The last two months of 2016 had estimated increases in unadjusted retail sales of 4.2 and 16.5 per cent.  In 2015, the two retail sales estimates were 0.9 and 17.2 per cent. November and December unadjusted US retail sales from 2014 were 0.8 and 15.2 per cent, respectively.  Continuing to go back in years, the estimates for 2013, 2012, and 2011; the November and December estimates were 2.4 and 13.7 per cent, 3.6 and 13.1 per cent, and 4.3 and 16.3 per cent.

Readers may have noticed that the sum of the two estimates from November and December of 2022 are clearly the lowest sum of all of the sums of November and December US retail sales estimates each year going back through 2011 (and farther back).  Simple numerical averages can be calculated by dividing each of these sums by two.  Given that these averages of the November and December percentage changes in retail sales provides an estimate (probably without compounding) of the average monthly growth rate of retail sales during these years, the November and December 2022 period clearly has the lowest average growth in US retail sales based on unadjusted data.  This may be why the seasonally adjusted data show declines in November and December of 2022.  Adding on to that but on an unadjusted basis the decline in January 2023, and a decline that was probably larger in real or inflation-adjusted terms, the economic fear and angst that many seem to be feeling becomes much easier to understand.

As an additional aside on seasonally adjusted versus unadjusted US retail sales, with the unadjusted series, of the first nine months of 2022, six of the nine months had reported declines.  Only March, May, and August of the first nine months of 2022 showed gains.  Realizing that from my calculations, the unadjusted CPI from the US Bureau of Labor Statistics (2023) increased in all but one of those months, with August 2022 the lone exception, the decreases in unadjusted real retail sales were more pronounced.

Readers may recall that my previous blog post with the short title “Falling Employment and Debt Deflation” (Hartman 2023), I pointed out that on an unadjusted basis, the estimates are that US employment actually fell in January 2023.  I also pointed out that second quarter employment in the US last year may have actually fallen by more than 270,000 jobs.

Again, readers should remember that millions are struggling to pay bills.  In addition to inflation and rising interest rates, economic inequality could be creating huge problems for many households.  Many have noted that paychecks have struggled to keep pace with inflation in the US for decades.  Unfortunately, the US may be in a period when at times fewer people are receiving paychecks due to job separations surging at times.  Keeping these factors in mind, the economic pain may be more palpable.

Is the January 2023 increase in the CPI just a short-term increase after decreases in November and December of 2022?  Or is it a return to surging inflation?  We’ll have to wait and see.

For now, though, many households probably have not noticed much, if any relief, in their economic plight.  The departure of Dr. Lael Brainard from the Federal Open Market Committee (FOMC) of the Board of Governors of the US Federal Reserve Bank could be particularly unwelcome news for millions in the United States.  Earlier today, Professor and former FOMC Vice Chair Alan Blinder discussed this development on CNBC with some regret.  Dr. Brainard helped to give a voice for millions when the FOMC was discussing future monetary policy decisions. 

As Dr. Brainard departs the FOMC and joins the Biden Administration, will these millions of people still have a voice?    It is important for monetary policy to control inflation, as inflation has eroded buying power when incomes have failed to keep pace with inflation. But would a wait and see approach for monetary policy be better now?  If tight monetary policy causes people to lose their jobs, then how can they continue paying bills?  Is it either fair or wise to cause people to lose jobs to reduce inflation, particularly when some suggest that inflation may be under control and so many have already been suffering for so long?  Can modern money theory help to (a) offer solutions and (b) mitigate the harmful effects of job losses?  Moreover, given that monetary policy usually if not always has an effect with a delay or a lag, is it better to pause increases in the federal funds rate if the unadjusted data may possibly be painting a picture of weaker aggregate demand in the economy than the seasonally adjusted data?

Please let me present one final thought for monetary policy and fiscal policy before concluding.  Much of this posting has dealt with seasonally adjusted data versus data unadjusted.  Clearly, seasonally adjusted data is essential for some purposes.  But, can people feel seasonally adjusted data?

 

I apologize in advance for any misstatements, misunderstanding or confusion.  Also, please realize that subsequent data revisions may change the above data and in turn the analysis.  Constructive comments are welcome at my Twitter account (@HarrisonCHartm1).  Please realize that I cannot respond individually to all comments.  Thanks in advance for your understanding

 

REFERENCES AND FURTHER READING

 

Blinder, A.  (2023)  Interview on CNBC with M. Santoli and K. Quintanilla,  February 15, 2023, available at:  https://www.cnbc.com/video/2023/02/15/what-lael-brainards-departure-means-for-the-fed.html

 

 

 

https://www.bls.gov/news.release/cpi.nr0.htm

 

U.S. Bureau of Labor Statistics. (2023) Consumer Price Index for All Urban Consumers: All Items in U.S. City Average [CPIAUCNS], retrieved from FRED, Federal Reserve Bank of St. Louis; https://fred.stlouisfed.org/series/CPIAUCNS, February 15, 2023.  Note that the data was accessed from:

https://fred.stlouisfed.org/data/CPIAUCNS.txt

 

Cox, J.  (2023)  “Retail Sales Jump 3% in January, Smashing Expectations Despite Inflation Increase.”  February 15, 2023, available online at:

https://www.cnbc.com/2023/02/15/retail-sales-january-2023-.html

 

 

Dickler, J. (2023)  “U.S. Credit Card Debt Jumps 18.5 Per Cent and Hits a Record $930.6 Billion.”  February 3, 2023, available online at:

https://www.cnbc.com/2023/02/03/us-credit-card-debt-jumps-18point5percent-and-hits-a-record-930point6-billion-.html

 

Hartman, H.C. (2023)  “Falling Employment and Debt Deflation:  Two Possible Problems Now and Beyond Valentine’s Day Due to Tight Monetary Policy.”  February 3, 2023, available online at:  https://harrisonhartman.blogspot.com/2023/02/falling-employment-and-debt-deflation.html

 

Hartman, H.C.  (2023)  “Will Food for Thought Feed the Fed while People Try to Feed Their Families?”  January 29, 2023, available online at https://harrisonhartman.blogspot.com/2023/01/will-food-for-thought-feed-fed-while.html

 

U.S. Census Bureau.  (2023)  Advance Retail Sales: Retail Trade [MARTSMPCSM44000USN], retrieved from FRED, Federal Reserve Bank of St. Louis; https://fred.stlouisfed.org/series/MARTSMPCSM44000USN, February 15, 2023.   Note that the data were accessed from:  https://fred.stlouisfed.org/data/MARTSMPCSM44000USN.txt

 

U.S. Census Bureau.  (2023)  Advance Retail Sales: Retail Trade [MARTSMPCSM44000USS], retrieved from FRED, Federal Reserve Bank of St. Louis; https://fred.stlouisfed.org/series/MARTSMPCSM44000USS, February 15, 2023.  Note that the data were accessed from:  https://fred.stlouisfed.org/data/MARTSMPCSM44000USS.txt

 

FALLING EMPLOYMENT AND DEBT DEFLATION: TWO POSSIBLE PROBLEMS NOW AND BEYOND VALENTINE’S DAY DUE TO TIGHT MONETARY POLICY

Have you heard that when looking at employment estimates in the United States for January 2023 not on a seasonally adjusted basis, but rather based on the raw unadjusted data, employment in the U.S. fell in January 2023?  This is true regardless of whether we look at (1) the household survey data which is seen as being a ‘noisier’ series more subject to monthly randomness based on the households that happen to be surveyed, or (2) the establishment data, which is in part based on estimates of the number of new businesses created and the number of firms that unfortunately go out of business.

While CNBC, MSNBC, and others trumpeted the estimated gain of 517,000 jobs from the establishment survey conducted by the US Bureau of Labor Statistics ,and I have been told that at least one person noted ‘on air’ that the household survey showed even stronger growth, both of those were seasonally adjusted estimates.  Using seasonally adjusted numbers from the St. Louis Federal Reserve Bank FRED website, I estimate that the household survey announcement today (Friday, February 3, 2023) for January 2023 reflects an increase of nearly 900,000 jobs!  But, in terms of actual numbers, my calculations show that nonfarm payroll employment not seasonally adjusted fell by more than 2.5 million jobs, while the household survey shows a more modest decline of about 180,000 jobs.  Those are both job losses, not job gains!

In the early afternoon or the very late morning on CNBC today, Steve Liesman noted that Mark Zandi of Moody’s advised people on Twitter (@Markzandi) that they may be better off ignoring the data release, perhaps due to difficulty in making seasonal adjustments. Earlier on CNBC Jan Hatzius, Chief Economist at Goldman Sachs, noted that every January shows large numbers of net job losses, making it difficult to determine appropriate seasonal adjustments.  In fact, in a similar vein, if my memory is correct, then every first quarter shows a loss of real GDP given that holiday shopping is so strong in November and December at the end of the previous year.  So, maybe this means that the job market in January was bad – but just not as bad as it usually is in January?!

Combined with the release last week on January 25 stating that in the second quarter of 2022, U.S. employment may have fallen by about 273,000 jobs, do these findings suggest that the U.S. job market may be softening?  If so, then did the Federal Open Market Committee (FOMC) of the Federal Reserve Bank (the Fed) consider that when raising its target for the federal funds rate?  My earlier blog post (short title “Will Food for Thought Feed the Fed”, see internet address in the References and Additional Reading section below) noted that I would have preferred that the FOMC would have kept the federal funds rate unchanged to wait for further information before making a change. 

I am concerned that both the Fed and the U.S. Bureau of Economic Analysis (BEA) may place too much emphasis on the seasonally adjusted estimates and not enough emphasis on the estimates not seasonally adjusted and also not enough emphasis on the January 25 revisions from the BLS.  The BEA may rely on the seasonally adjusted estimates for employment in estimating the growth of first quarter real GDP in 2023 in the U.S.  But what if some of the decrease in actual jobs in the US economy in January 2023 is not due only to seasonal factors but due to an economic downturn?  The BEA may not detect this as soon as possible if it relies too much on the seasonally adjusted estimates.  Regardless of what the BEA does, tighter monetary policy could very easily lead to a recession with a loss of jobs and real GDP by making it harder to borrow funds and by taking money out of the economy, discouraging consumption by households and investment in physical capital by businesses.

Another concern that I have about the approach by the FOMC in reducing inflation in the United States deals with the emphasis on calculating inflation based on a measure of the price level in the most recent period compared with one year ago or even six months ago rather than compared with the most recent month or quarter.  Wharton School University of Pennsylvania Professor Jeremy Siegel and likely others have noted that at least some measures of US inflation measured month to month have gone negative, implying that on average, those weighted averages of prices in the US economy have started to fall.  The US economy may have a situation now where some prices in the economy have fallen in recent months but still could be higher than they were a year ago or six months ago.

A potential problem with the FOMC focusing on inflation data with inflation calculated based on prices six months ago or twelve months ago is that in order for the FOMC to reach its stated target of two per cent on a ‘year over year’ basis, given that inflation has been noticeably higher than two per cent, weighted averages of prices must decrease in the short term more than they have already.  This could lead to a problem known in economics as debt deflation. 

Wages and salaries may drop as part of the deflation.  When borrowers have fixed nominal interest rate loans and prices in the economy drop, the borrowers have a greater sacrifice of purchasing power required to repay loans.  In the case of the US for example, borrowers have to repay loans in dollars worth more in buying power than they anticipated (unless they were expecting a level of deflation at least as large as the one consistent with the level of deflation in the economy).  Further, borrowers’ wages and salaries may fall, yet with fixed nominal interest rate loans, they may not get a break in terms of the nominal amount that they must pay when they repay loans.  With price levels in the economy falling but fixed interest rate loan repayments remaining unchanged, inflation-adjusted or real loan repayments go up.

Because borrowers may reduce their spending to repay loans more than lenders increase their spending as they are repaid (for example, see Krugman and Wells (2009, p. 463)), real GDP would almost certainly be lost, and unemployment would almost surely rise.  In addition to the greater hardship associated with repaying because incomes may be falling, borrowers would have a greater strategic incentive to default than without deflation due to the need to give up more purchasing power to repay loans, so that a financial crisis could arise due to a surge in defaults.  Given that credit card debt may already be at record-setting levels and delinquencies may be surging, would the economy be better off with the FOMC of the Fed switching to more of a ‘wait and see’ approach rather than its relatively new fixation on reducing inflation, especially when inflation may already be falling?

We can hope that the US job market is as strong as the seasonally adjusted estimates imply.  However, my concern is that with millions of households struggling to pay bills, unfortunately actual labor market conditions and tight monetary policy may have many broken hearted before, on, and after Valentine’s Day.

 

Readers may enjoy an earlier blog post that I wrote dealing at least in part with the household employment estimates versus the establishment survey and seasonally adjusted data versus data not seasonally adjusted.  A short title could be “Is a Temporary Dichotomy Emerging Between…” and a website address for this entry can be found in the References and Additional Reading section below.

Please note that subsequent data revisions may change the analysis in the blog entry above, and that data were accessed from the FRED website of the St. Louis Federal Reserve Bank.  Constructive comments are welcome at my Twitter account (@HarrisonCHartm1).  Also, please accept my apology for any misstatements, inconvenience, and/or confusion.

 

 

REFERENCES AND ADDITIONAL READING

Hartman, H. C.  (2020)  “Is a Temporary Dichotomy Emerging Between Establishment Data and Household Data Regarding U.S. Employment?”  March 7, 2020, available online at https://harrisonhartman.blogspot.com/2020/03/is-temporary-dichotomy-emerging-between.html

Hartman, H. C.  (2023)  “Will Food for Thought Feed the Fed While People Try to Feed Their Families?”  January 29, 2023, available online at https://harrisonhartman.blogspot.com/2023/01/will-food-for-thought-feed-fed-while.html

Krugman, P. and Wells, R.  (2009) Macroeconomics, 2^nd edition,  New York, New York, Worth Publishing, p. 463.

Tweets of Mark Zandi (@Markzandi), February 3, 2023.

U.S. Bureau of Labor Statistics, Employment Level [LNU02000000], retrieved from FRED, Federal Reserve Bank of St. Louis; https://fred.stlouisfed.org/series/LNU02000000, February 3, 2023.

U.S. Bureau of Labor Statistics, All Employees, Total Nonfarm [PAYNSA], retrieved from FRED, Federal Reserve Bank of St. Louis; https://fred.stlouisfed.org/series/PAYNSA, February 3, 2023.

U.S. Bureau of Labor Statistics, Employment Level [CE16OV], retrieved from FRED, Federal Reserve Bank of St. Louis; https://fred.stlouisfed.org/series/CE16OV, February 3, 2023.

U.S. Bureau of Labor Statistics, All Employees, Total Nonfarm [PAYEMS], retrieved from FRED, Federal Reserve Bank of St. Louis; https://fred.stlouisfed.org/series/PAYEMS, February 3, 2023.