Why You Can’t Put Faith in Reports of First-Quarter Economic Slumps

This is the same conundrum that affects the daily explanation of why the stock market exchanges went up or down. News folks must have a reason so they invent one as if it represents some sort of truth. So, yesterday, the markets were down because of the lack of consumer spending. Well of course consumer spending is down, and not just in the first quarter, because companies stopped paying a living wage to those who must spend their wages to survive. And so it goes ...

A senior econ major where I teach [Washington & Lee University] just did a capstone paper on seasonality, which shows up in (un)employment data and many other output series (but not price series). It is also not unique to the US, and not unique to the Northern hemisphere. His observation (based primarily on a literature survey) is that the larger first quarter effect reflects a trend, consistent with what you write about. Seasonality corrections have a hard time dealing with that. The longer a trend persists the easier it is build it into the seasonality ARIMA model. To the extent that the pattern is strengthening, year-on-year comparisons won't fully correct for it, either.

Now the origin of the pper topic stemmed from the room where the class was held, in which H. Parker Willis had his office a century-plus ago. Willis' work for Senator Glass on a central bank proposal became the core of the Federal Reserve legislation. And one impetus for creating the Federal Reserve was to offset the strong seasonality in interest rates stemming from the still-large role of agriculture. In mentioning the import of the room at the start of the term I wondered aloud what had happened to seasonality, and got a nice paper!!

Present USA economic policies are only creating a larger population of poor, increased economic control by a minority of the rich and creating a fast disappearing middle class America.
Present economic policies are rapidly turning the United States into an aristocracy. Economic policies in this article are only "theoretical economies". The hard reality is these "fantasy land economic policies" demand much of its working population, for they are the ones that pay the price, living reality in a "fantasy land" economy, demanding its workers to face daily employment mostly in low wage, jobs that are only more and more oriented to the service jobs industry. Workers under these economic policies will only continue face a reality of massive exportation of jobs, a rapidly declining manufacturing industry and finding only mostly low pay jobs. Present economic policies need to have built in job safety nets for the working people. Sad day when the United States of America cannot longer support dreams...the "All American Working Class Dream".

Good phenomenon to know about. I suspect it has something to do with the massive consumer holiday spending in Q4, much of it on imported goods. That inventory starts arriving as early as Sept and is pushed out in increasingly desperate discounting before and after the holidays, especially Black Friday after Thanksgiving. I don't recall this Q4 bunching of retail activity to the same degree in the 1970s or before, when the average personal savings rate peaked at 12% on much lower real income, before reaching a low of 2% around 2006, now around 5%.

http://m.research.stlouisfed.org/fred/series.php?sid=PSAVERT&show=ch...

Perhaps there are other forces at work here that I am not familiar with, however, if 70% of U.S. GDP is based on consumer spending, yet, numbers show the workforce has shrunk approximately 2% since 2007 AND wages have relatively stagnated, the quoted unemployment rate along with all this discussion about the significant rebounding of the economy, to me, just doesn't add up.

GDP numbers don't reflect reality. Statistical manipulation
and methodology can change the numbers drastically.
Recently India changed methodology, its GDP jumped
from 5% to 7% without significant change in the economy.
Nigeria changed its methodology and its GDP doubled.
USA also has revised methodology occasionally with
positive net effect. Recently Britian revised its methodology
to include income from prostitution which spiked its
GDP. As a single number GDP is not that important,
all other indicators like employment, retail sales,inflation,
debt,industrial production, imports/exports give better
picture of the economy. In USA consumer spending account
about 70% of GDP, personal income, debt and retail sales
tell more about the economy than the fluctuations in the
inventory.

Non-farm business productivity is rising at the slowest rate in 50 years. And the velocity of money – the speed at which each unit of currency changes hands and a key component of inflation – has fallen to the lowest level in half a century. Why?

Because of the declining marginal utility of debt. When there is little debt, you can add cash and credit to a system and get a boost. The money circulates. The economy revs up. But the more you add, the greater the burden of debt becomes… and the less of a boost you get from it.

Finally, you’ve ballooned the Fed’s balance sheet to $4.5 trillion and you’re getting a measly 1% per capita GDP growth in return. And then… as in the first quarter of this year… the growth falls to near zero. All the “stimulus” since 2000 was a scam. It stimulated nothing but more debt – which slows the rate of real growth.

Sales in my retail store are down and sales in many established retail stores were down in March (Macy's - 1.66%, JC Penny -2.66%, Kohls -1.48%, WalMart -0.18%) or anemic (Target + 0.12%) compared to a year ago (CNN 4/17/2015 Statistics within parentheses). That's the truth. The economy is actually not recovering from a retail sales perspective.

Government numbers are never to be trusted, in addition to any economic forecast. Many make the numbers up anyway and are used primarily for government expediency.

Are there residual seasonal patterns for the summer months as well?
If that exists too, perhaps this is a measure of the effects of global warming. The weather patterns started to go out off long-term patterns back in the 1990s (exactly when your economic skew started).

One way to avoid the issue of seasonal adjustment is to look at the percentage change in GDP from the year-ago quarter. The result shows that GDP growth in the current recovery has been fairly consistent at around 2% to 2.5%. This is even more obvious if inventory changes are subtracted yielding real final sales of domestic product. See chart:

http://research.stlouisfed.org/fred2/graph/fredgraph.png?g=18YB

What's remarkable is how steady real final sales growth has been in this recovery. The highest yearly reading came last year in the Q3 at 2.75% but the media at the time only heralded the 5% quarterly GDP figure which merely reflected the bounce back from the 2.1% contraction in Q1 of last year.

The main point here is that the change from the prior year's quarter is always calculated by the BEA in their GDP reports but the media never bothers to inform the public preferring only to report the quarterly data which is subject to possibly erroneous seasonal adjustment. Then when the Q1 result is dismal they blame the weather but when the bounce comes in the later quarters the interpretation is that the economy is finally accelerating. The media also do not report on the final sales data which removes volatile inventory changes.

This phenomenon does not show up in the Philadelphia Fed measure GDPplus (http://bit.ly/1INTIbE), which uses data from multiple sources to reduce errors in measuring GDP growth. The average growth rates by quarter and decade using GDPplus are:

1960's: Q1 = 4.83, Q2 = 4.26, Q3 = 4.01, Q4 = 4.04
1970's: Q1 = 3.00, Q2 = 3.17, Q3 = 3.14, Q4 = 2.76
1980's: Q1 = 3.22, Q2 = 2.65, Q3 = 3.32, Q4 = 3.41
1990's: Q1 = 3.27, Q2 = 3.61, Q3 = 3.31, Q4 = 3.58
2000's: Q1 = 2.29, Q2 = 1.68, Q3 = 1.41, Q4 = 1.24
2010's: Q1 = 2.20, Q2 = 2.49, Q3 = 2.91, Q4 = 2.61

So it looks to me like the strange Q1 results are just measurement error, not actual movements in GDP, although it would be interesting to know why this is going on.

Has something changed in the calculations? Is there something being done that's different? Are companies reporting things differently than they used to? Is it because of the Christmas holiday? Or is it just one of those things?

Please consider the following three sources of variability: 1) The December inventory overhang or lack thereof for retailers may significantly impact the GDP growth rate for the 1st quarter. As the US macroeconomy has moved into a greater service sector mode of operation, service sector production has tended to overshadow manufacturing sector production. But, in general, service sector production cannot be stored; whereas goods inventories can be stored. Given this longer shelf-life of goods as opposed to services, mistiming of goods' production can affect future production. Simply put, with about 40% of many retailers sales compressed into the December holiday season, this inventory hangover effect can materially impact 1st quarter production. 2)Personal income tax refunds also are compressed into the 1st quarter of the year. These refunds can impact consumer spending in the 1st quarter. Also, income tax law changes, for example, passage of the ACA and its tax implications, can impact taxpayer refunds. Yet, I think, the balance of personal income tax liability payments can generally be deferred, without penalty until April, or the 2nd quarter, if the taxpayer meets certain minimal payment criteria, such as having paid the IRS 100% of the prior year's tax liability by mid-January, or 90% of the current year's tax liability by then. 3) Wall Street bonuses are also paid in the 1st quarter and can be quite volatile, as the results from many trading markets are. Th 4/23 1:38 p

Would make a much more interesting article if the author had shown fed predictions, or any leading economist predictions as compared to actual performance. Far to much talk in our world about predictions when the reality is past predictors are allowed to pontificate ad nauseam when they are in fact poor predictors.

So I guess we're really doing great and Obama saved the economy after all.
Thanks, Obama.
You can go now.

The author of this article notes this residual seasonality was not present in the 70s. Look at the chart. Look how this has crept towards its current trend. Wages have stagnated since the 70s. Every year people are spending money they don't have in the last quarter (using credit cards to fund holidays) and tightening in January- March when they come to their senses. These periods are so intense in their effort to pay the card back down there is no way to mitigate it through massaging data- seasonally adjusting it.

From the article the government employee states they were “currently examining residual seasonality in several series..." The author notes he could find none in other indicators and sites gross domestic income. There is no need to look at gross domestic income. The lack of growth in the first quarter of GDP does not indicate people are making less than usual- they just got a sobering reminder when they opened their credit card bill.

Maybe we should be questioning fourth quarter GDP. Maybe we should measure how much of that GDP number is a reflection of Americans going into debt to live like they did in the 70s.

Perhaps Northern Hemisphere economists should adopt the laypersons name for the the First Quarter, "winter," and just pick things back up in April like most of the rest of us.

I saw the Liesman report on this topic yesterday. Superb reporting and analysis.

Reality check lets look close whats the total utization of present manufacturing in usa factorys/ . Whats total population that can work an those with out job ? Whats total jobs make up recovery are part time or under 32 hours week? Whats total debt government spends on imports?

Maybe it is time to find another indicator to use as the primary figure of merit for economic performance.

A genuinely interesting trend but, please, GDP is already highly "massaged" data.

GDP gets variously trotted out by wall streeters, Fed Bank. and Harvard economists whenever convenient to announce it to the press. The spinfactor on GPD is notable.

Every economic statistic published by government is "seasonally adjusted". In other words, a fudge factor is applied to annualize rate calculations.

One could compare the present quarter to the same quarter a year ago without fudge factors, but that isn't how they do it.

If the situation is different, namely strong 1st quarter every year, I wonder if similar question comes out. China has been doing this, not only the 1st quarter, but also every quarter and every year. Even current prime minister of China said once, 'GDP number is man-made'. We are lucky that our GDP data is more trust worthy.

Yet another attempt to excuse away the failed economic policies of this administration. Seven years of 0% interest have distorted markets, harmed retirees and fixed income investors, and thrown unknown numbers out of work by lowering the cost of capital as compared to the cost of labor. Now, these same people are pushing a $12/hour minimum wage, which will harm low skilled workers even more.

As I read it, this column is taking the position that economic reports are to be believed only insofar as they are favorable to the administration. Otherwise, they are to be ignored.

Once upon a time there was a drunken Texan. He was very, very drunk. He staggered out behind the saloon where there was a barn. He fired his pistol at the barn hitting it most of the time. Of course the bullet holes were perfectly random. Being perfectly random they were not evenly spaced, but tended to bunch up in places. The next morning the Texan woke up and looked at the barn. He the took some paint and painted a bulls eye pattern around each of the bunches. He then showed the barn to his friends and told them this proved that he was an accurate shot even when dead drunk.

Why not just call it what we, when we were part of corporate America, called it," the year end hangover". American business is based on growth, just not first quarter growth. The last quarter of the year is where most CEOs and top management make their bonuses. They often work to push business into a current fiscal year, to worry about next year's numbers later. This often becomes a vicious cycle with the need for stellar year end performance stressing organizations to the point that they struggle to maintain their performance levels into Q1. You want to predict the 1st Quarter, look at the prior Q4. This is not a secret and does not require a Harvard MBA to know about.

Sorry for sounding so critical or conspiratorial, but the thrust of the argument seems to be part of a larger school of economic apologists. We've had absolutely atrocious macro-economic data recently. If you combine that with a widely-accepted view that the unemployment numbers are generally useless, articles such as these are merely part of some narrative that "everything is great" and such high valuations for stocks is justified. The counter argument is that 3 years does not a data set make, and the economy's numbers are bad for a reason: the underlying economy is terrible, well except for the S/p 500 and the NASDAQ.

So articles such as this, parroting the talking heads of CNBC is to merely support political agendas.

A minor factor in this phenomenon is possibly, incredibly enough, campaign spending. There is an obscure series in consumer spending called "Professional advocacy, gross output" which includes this stuff. Not surprisingly, it has a pattern of collapsing in the first quarter of odd-numbered years, especially after those divisible by 4. Not something that seasonal adjustment programs can handle that well. On average over the last 15 years there's a 10 percentage point downshift in the annual rate of growth in this series between the 4th and 1st quarter (and that's taking into account the acceleration at the start of even-numbered years). It's a small series (no more on average than two-tenths of one percent of GDP), so it's not by any means the big player in this but at least it's systematic. Real consumer spending in the first quarter was, judging by the monthly data through February, held down around $10 billion by this, which would hold down the annual rate of overall GDP growth by a couple of tenths of one percent.
The impact on consumer spending would naturally be somewhat greater than that on overall GDP, so this a part of an explanation for the puzzle as to why the reported real consumer spending series have seemed lackluster given the plunge in gas prices.