HomeLatest ThreadsGreatest ThreadsForums & GroupsMy SubscriptionsMy Posts
DU Home » Latest Threads » progree » Journal
Page: 1

progree

Profile Information

Gender: Male
Hometown: Minnesota
Member since: Sat Jan 1, 2005, 03:45 AM
Number of posts: 5,421

Journal Archives

EF-0. Economic Statistics with links to official sources

Major updates were done to all pages on February 2, 2019. For additional updates that were done since then, please see "Recent Topic Updates" at the bottom of this EF-0 page. (the quickest way to get there might be to click on this link: http://www.democraticunderground.com/111622439#post1 AND THEN SCROLL UPWARD.

I won't be doing monthly updates, but rather annual updates of the jobs number, so if the job statistics are somewhat out of date in the future, I hope people will read "Beware the tricks of the economic pundits out there". The other pages (EF-1 through EF-10) also has information that is still relevant or relatively timeless. I also make occasional changes to some of the other pages, particularly the debt and deficit information on the EF-5 page.

Almost all sections have where to find the official numbers, such as the Bureau of Labor Statistics and Treasury.gov, or widely trusted non-partisan sources. Hopefully people will find these pages a useful reference for finding information.

There are some amazingly distorted presentations of what their numbers are and what they mean that you find on the web, and yes, DU too. Please see the "Beware the tricks of the economic pundits out there" section in the bottom half of this page for examples of what I mean by distorted presentations of BLS statistics.

I don't claim that the BLS and other government sources are inerrant, or even unbiased, e.g. whoever came up with some of the definitions like the official (U-3) unemployment rate being a count of jobless people who looked for work sometime in the past 4 weeks. And it is obvious that most of the Household Survey numbers have a lot of statistical error, considering how they wildly bounce around from month to month.

I'm just saying all of this is a presentation of the actual BLS numbers (for the most part) with links to the statistics being discussed, so that you can check it out for yourself.

Unfortunately, a lot of the formatting has been lost thanks to the May 2017 hack of the DU website. For the latest version at archive.org -- WHICH SHOWS THE ORIGINAL FORMATTING -- see:

https://web.archive.org/web/20160411173355/http://www.democraticunderground.com/111622439

Unfortunately the latest archive.org snapshot that shows the original formatting is April 2016. Oh well. There is a way to ask archive.org to save a current snapshot ... the "Save Page Now" feature at https://archive.org/web. But unfortunately any snapshots made after the May 2017 hack of the DU website appear the same as what you are looking at -- with only selected formatting restored.

For more on what formatting does and does not work at DU, see https://www.democraticunderground.com/?com=view_post&forum=1256&pid=13211

Here is a table of contents of this thread:

{#} EF-1. Job Loss and Creation - Payroll Employment. At the bottom all post-WWII presidents with completed terms are compared

{#} EF-2. Unemployment Rate, Labor Force Participation Rate, Unemployment Insurance Claims

{#} EF-3. Recessions and Expansions - Official (NBER.org). Also GDP (Gross Domestic Product)

{#} EF-4. U.S. Stock Market as measured by the S&P 500 and the Dow Jones Industrial Avg

{#} EF-5. National Debt. Budget Deficits and Surpluses

{#} EF-6. U.S. Dollar Index (DXY). Oil Prices

{#} EF-7. In Progress (mostly Dem presidencies v. Repub presidencies. Also Inequality)

{#} EF-8. In Progress - Some canned excerpts to use in the message board wars

{#} EF-9. Incomes and Inequality and Consumer Prices and Poverty (in progress)

{#) EF-10. Definitions, Links (In Progress)

I use facts from these in mixed message boards and in comments on news articles such as at news.yahoo.com. Be aware that I have included a few statistics that are not so pleasant as far as Obama's record, ones that anyone debating with others should be aware of because occasionally you will see these points or they will come back at you with these statistics (forewarned is forearmed).

Here's another major major economy resource: CabCurious' "Factual talking points on the economy" - lots of very interesting graphs - take a "scroll" through them. But it hasn't been updated since fall of 2012. http://www.democraticunderground.com/125170175

##################################################################################
##################################################################################

Here are some summary tables of the key jobs reports statistics from the Establishment Survey and the Household Survey released on February 1, 2019.

A narrative "Detailed Discussion" section follows these tables.


In the below tables, all "%" ones are percentage point changes, *not* percent increases or decreases. FOR EXAMPLE, when you see something like this:

+0.1% Unemployment rate

It means that the unemployment rate increased by 0.1 percentage points (this EXAMPLE is from March 2016 when the unemployment rate rose from 4.9% to 5.0%).

Before each item, (F) indicates very bad, (D) indicates bad, (C) indicates neutral, (B) indicates good, (A) indicates very good. (Edit: I haven't been putting these "grades" on the items lately)

PLEASE NOTE THAT ALL THE NUMBERS IN THE TABLES BELOW ARE SEASONALLY ADJUSTED

See "Detailed Discussion" section below for a narrative discussion of the above statistics over the past month, the past year, and since the jobs recovery began in March 2010.

The links to the data below in the "over the last year" etc. tables
# Nonfarm Employment (Establishment Survey, https://data.bls.gov/timeseries/CES0000000001
# INFLATION ADJUSTED Weekly Earnings of Production and Non-Supervisory Workers http://data.bls.gov/timeseries/CES0500000031
# Labor Force http://data.bls.gov/timeseries/LNS11000000
# Employed http://data.bls.gov/timeseries/LNS12000000
# Unemployed http://data.bls.gov/timeseries/LNS13000000
# ETPR (Employment-To-Population Ratio) aka Employment Rate http://data.bls.gov/timeseries/LNS12300000
# LFPR (Labor Force Participation rate) http://data.bls.gov/timeseries/LNS11300000
# Unemployment rate http://data.bls.gov/timeseries/LNS14000000
# U-6 unemployment rate http://data.bls.gov/timeseries/LNS13327709
# NILF -- Not in Labor Forcehttp://data.bls.gov/timeseries/LNS15000000
# NILF-WJ -- Not in Labor Force, Wants Job http://data.bls.gov/timeseries/LNS15026639
# Part-Time Workers who want Full-Time Jobs (Table A-8's Part-Time For Economic Reasons) http://data.bls.gov/timeseries/LNS12032194
# Part-Time Workers (Table A-9) http://data.bls.gov/timeseries/LNS12600000
# Full-Time Workers (Table A-9) http://data.bls.gov/timeseries/LNS12500000
# Multiple Jobholders as a Percent of Employed (Table A-9) https://data.bls.gov/timeseries/LNS12026620

# Civilian non-institutional population https://data.bls.gov/timeseries/LNS10000000

Some statistics by age group
The ones beginning with "LNS" are seasonally adjusted (SA).
The ones beginning with "LNU" are not seasonally adjusted (NSA)
16+ is the default one that the BLS and the media report.
16+ means every civilian non-institutionalized person age 16 and over, including centenarians. So it is misleading -- the population is aging and there are about 10,000 boomer retirements a day (which comes to 3.6 million/year). That's why I show other age groups.

Age 25-54 is what the BLS calls the "prime age". It isn't contaminated by a lot of voluntary retirements.

By default, the graphs are 10 - 11 years, specifically they begin in the January of the year that was 10 years ago -- meaning in this case they begin January 2009 (which happens to be near the bottom of Great Recession job market -- well actually the job count fell for another 13 months to its lowest point in February 2010)

You might want to set the start date of the calendar back to, oh, whatever. 1989? 1979? 1969? In order to get a more historic view. 1989 is about when the rapid growth of female workforce participation began to level off

LFPR - Labor Force Participation Rate for some age groups
The LFPR is the Employed + jobless people who have looked for work in the last 4 weeks (and say they want a job and are able to take one if offered). All divided by the civilian non-institutional population age 16+.
SA means Seasonally adjusted. NSA means Not Seasonally Adjusted
16+: SA: LNS11300000 NSA: LNU01300000
25-34: SA: LNS11300089 NSA: LNU01300089
25-54: SA: LNS11300060 NSA: LNU01300060
55+: SA: LNS11324230 NSA: LNU01324230
65+: SA: ---------------- NSA: LNU01300097

ETPR - Employment to Population Ratio for some age groups
SA means Seasonally adjusted. NSA means Not Seasonally Adjusted
16+: SA: LNS12300000 NSA: LNU02300000
25-54: SA: LNS12300060 NSA: LNU02300060
55+: SA: LNS12324230 NSA: LNU02324230
65+: SA: ---------------- NSA: LNU02300097


OVER THE LAST YEAR (last 12 months), i.e. Trump's 2nd year:
Updated to include the January 2019 jobs report that came out on Feb. 1, 2019
==== ESTABLISHMENT SURVEY ====
+2,807,000 Nonfarm Payroll Employment (Establishment Survey, CES0000000001)
+2.37% INFLATION ADJUSTED Weekly Earnings of Production and Non-Supervisory Workers ( CES0500000031 )
......... the weekly earnings percentage is 11 months thru December because no CPI data for January yet
==== HOUSEHOLD SURVEY ========
+2,106,000 Labor Force = Employed + jobless people who looked for work in the past 4 weeks
+2,212,000 Employed
-106,000 Unemployed (jobless people who looked for work in the past 4 weeks)
+0.5% Employment-To-Population Ratio aka Employment Rate
+0.5% LFPR (Labor Force Participation rate)
-0.1% Unemployment rate
-0.1% U-6 unemployment rate (fabulous. it includes anyone that looked for work even once in the past year)
-0.04% "U-7" unemployment rate: Counts EVERY jobless person who SAYS they want a job,
` ` ` no matter how long it has been since they looked for work, plus part-timers who want
` ` ` full time work
+89,000 Not in Labor Force, Wants Job LNS15026639
+165,000 Part-Time Workers who want Full-Time Jobs (Table A-8's Part-Time For Economic Reasons)
-543,000 Part-Time Workers (Table A-9)
+2,792,000 Full-Time Workers (Table A-9)

The "U-7" unemployment rate is a creation of Paul Solman of the PBS Newshour, not a BLS number. The above number is one I calculated, because he doesn't update his number every month, and when he does, it is about a day after the jobs report comes out. My number has consistently matched his within 0.1 percentage points (and mine has always been a bit higher). The "U-7" unemployment rate counts EVERY jobless person who SAYS they want a job, no matter how long it has been since they looked for work, plus part-timers who want full time work

For more background on the U-7 number, see: "If you count everyone who says they want a job, even if they have made no effort to find one in many years" at http://www.democraticunderground.com/111622439#post2


OVER THE LAST 2 YEARS (last 24 months), i.e. Trump's presidency:
Updated to include the January 2019 jobs report that came out on Feb. 1, 2019
==== ESTABLISHMENT SURVEY ====
+4,879,000 Nonfarm Payroll Employment (Establishment Survey, CES0000000001)
+2.63% INFLATION ADJUSTED Weekly Earnings of Production and Non-Supervisory Workers ( CES0500000031 )
......... the weekly earnings percentage is 23 months thru December because no CPI data for January yet
==== HOUSEHOLD SURVEY ========
+3,536,000 Labor Force = Employed + jobless people who looked for work in the past 4 weeks
+4,566,000 Employed
-1,030,000 Unemployed (jobless people who looked for work in the past 4 weeks)
+0.8% Employment-To-Population Ratio aka Employment Rate
+0.3% LFPR (Labor Force Participation rate)
-0.7% Unemployment rate
-1.2% U-6 unemployment rate (fabulous. it includes anyone that looked for work even once in the past year)
-1.5% "U-7" unemployment rate: Counts EVERY jobless person who SAYS they want a job,
` ` ` no matter how long it has been since they looked for work, plus part-timers who want
` ` ` full time work
-481,000 Not in Labor Force, Wants Job LNS15026639
-606,000 Part-Time Workers who want Full-Time Jobs (Table A-8's Part-Time For Economic Reasons)
-633,000 Part-Time Workers (Table A-9)
+5,246,000 Full-Time Workers (Table A-9)

The reason there's no data for January yet for the inflation-adjusted Weekly Earnings is because the CPI inflation adjustment number for January is not yet available.

Most of the "over the last TWO years" numbers are really good numbers (the Trump bubble building?). Exceptions:

The Labor Force Participation Rate has edged up 0.3% over the past 2 years, to 63.2%, and is now only 0.8% above the 62.4% multi-decade low of September 2015.

Most of the "over the last year" numbers are good or very good numbers, but there are some rather bland ones too, indicating a considerable loss of momentum:

There is a big loss of momentum in bringing down the unemployment rate over the last year -- it declined by only 0.1 percentage points, from 4.1% to 4.0%. Likewise the broader U-6 unemployment rate has declined by 0.1 percentage points, from 8.2% to 8.1%. And the broadest measure of unemployment, Paul Solman's U-7 unemployment rate -- which simply asks jobless people if they want a job and counts them if they say yes -- is flat at 10.1%.

Over the past year, the number of people Not In The Labor Force who say they want a job has increased by 89,000. (This category of people are those who are jobless and have not looked for work in the past 4 weeks but say they want a job).

Part time workers who want a full-time job increased by 165,000 over the past year.


On a separate topic, in general, it seems to me that there is too much discussion in the media of the Labor Force Participation Rate -- aka the Labor Force to Population Ratio -- (the employed plus the jobless people who have looked for work in the last 4 weeks, all divided by the population), and not enough attention to what seemingly matters more -- the Employment to Population Ratio. Why aren't we highlighting the increase in the percentage of the population that is employed (the employment to population ratio)-- a figure that has been slowly moving up since the job market bottom, despite the growing wave of baby boomer retirements?

This was particularly irritating during the 2nd term of the Obama administration when the labor force participation rate was barely off its Great Recession bottom, while the employment to population ratio was definitely and steadily increasing since fall 2013.

(As always, the population being talked about is the civilian non-institutional population age 16 and over, including the elderly, even centenarians).

Part-Time Workers Who Want Full Time Jobs, as % of All Employed
Jan'17 Jan'18 Jan'19
3.8% ` 3.2% ` 3.3%

SINCE THE PAYROLL EMPLOYMENT RECOVERY BEGAN -- Last 107 months thru January 31, 2019: 1'19 - 2'10:

(This is the period from when continuous growth of payroll employment began, thru January 31, 2019)
Yes, they are all (with one exception) super-fantastic numbers in every way, which isn't surprising given that we are measuring from the very bottom of the Great Recession's job market. I think I will drop this report in favor of doing something more useful, like a comparison to, something like December 2003 which I have decided is mid-way between the dotcom bust and the housing bubble highs.

I've always been reluctant to compare to unsustainable super-peaks like the dotcom peak or the housing bubble peak because, well, they were manic and obviously unsustainable. And like I say, comparing to the very bottom is pretty useless after awhile.

One exception to the good numbers since the Great Recession jobs bottom: the Labor Force Participation Rate (LFPR) dropped 1.7%. Remember, we are talking about a drop of 1.7 percentage points from the very bottom of the Great Recessions's job bust (which was February 2010). Very peculiar. See for yourself: https://data.bls.gov/timeseries/LNS11000000

In contrast, the Employment To Population Ratio (ETPR) increased by 2.2% over the same period (February 2010 to January 2019). https://data.bls.gov/timeseries/LNS12300000 . This isn't surprising, it's what one would expect starting from the Great Recession's job bottom. What IS surprising is that LFPR and ETPR went in opposite directions, and both by sizable amounts.


Aren't most of the new jobs part-time?

No. This excellent post from early July 2015 show two perspectives of the trends in part-time workers and full-time workers (not part-time jobs and full-time jobs). Thanks mahatmakanejeeves
http://www.democraticunderground.com/10141134306#post12

Since February 2010 through January 2019, part-time workers have DEcreased by 875,000 while full-time workers have INcreased by 19,059,000. (Table A-9).

Chart 7 of the below link shows Part-time workers as a percent of total employed, Seasonally adjusted, 19902019. In recent years it has ranged from 20.1% at the height of the Great Recession to around 17% now.
https://www.bls.gov/web/empsit/cps_charts.pdf
https://data.bls.gov/timeseries/LNS12692153



What kind of Wages?

INFLATION-ADJUSTED Average WEEKLY Earnings Of Production And Nonsupervisory Employees, Total Private, 1982-84 Dollars
http://data.bls.gov/timeseries/CES0500000031 ## (Hourly earnings: http://data.bls.gov/timeseries/CES0500000008 )
#=#=#=>
Unfortunately, the graph has expired. But the key thing is that, as of February 1, 2019, the INFLATION-ADJUSTED weekly earnings of production and non-supervisory workers is up 2.6% over the last 2 years, and up 7.48% since February 2010 (the jobs market bottom)

Here is the nominal, i.e. not-inflation-adjusted version of the above:
Weekly: http://data.bls.gov/timeseries/CES0500000030
Hourly: http://data.bls.gov/timeseries/CES0500000008




See "Detailed Discussion" section below for a narrative discussion of the above statistics over the past year and the past two years (the past two years coincides with the Trump presidency


########################################################################
FFI on the most recent jobs report, straight from the Bureau of Labor Statistics: http://www.bls.gov/news.release/empsit.nr0.htm

Table A-1. Employment status of the civilian population by sex and age (household survey) http://www.bls.gov/news.release/empsit.t01.htm

Several graphs of the key economic stats -- http://www.bls.gov/web/empsit/cps_charts.pdf
Table of Contents as of 1/31/19
1. Civilian labor force
2. Civilian labor force participation rate
3. Civilian employment
4. Employment-population ratio
5. Nonagricultural wage and salary employment
6. Nonagricultural self-employed, unincorporated
7. Part-time workers as a percent of total employed
8. Employed part time for economic reasons
9. Civilian unemployment
10. Civilian unemployment rate
11. Duration of unemployment
12. Long-term unemployed as a percent of total unemployed
13. Reasons for unemployment
14. Job losers by layoff status
15. Unemployment rates for adult men, adult women, and teenagers
16. Unemployment rates by race and Hispanic or Latino ethnicity
17. Unemployment rates for persons 25 years and older by educational
attainment
18. Persons not in the labor force who want a job
19. Persons not in the labor force, selected indicators
20. Alternative measures of labor underutilization


The whole enchilada -- including all 16 "A" tables (the household survey) and all 9 "B" tables (the establishment survey) http://www.bls.gov/news.release/pdf/empsit.pdf

[font color = brown] ----------------------------------------------------------------------
Table A-1 and other tables can be found at the all-tables full jobs report at http://www.bls.gov/news.release/pdf/empsit.pdf, or gotten one-at-a-time from the bottom section of http://www.bls.gov/news.release/empsit.nr0.htm . For example, Table A-9 alone is at http://www.bls.gov/news.release/empsit.t09.htm )
----------------------------------------------------------------------[/font]

BLS Commissioner's Statement on The Employment Situation http://www.bls.gov/news.release/jec.nr0.htm

The Council of Economic Advisors' Take on the Jobs Report
https://www.whitehouse.gov/issues/economy-jobs/ (find this at
http://www.whitehouse.gov/administration/eop/cea )
The Council of Economic Advisors is a Trump admin propaganda organ now. Do not confuse it with the BLS (Bureau of Labor Statistics) which is supposed to be a non-partisan statistics-gathering government agency. On the other hand, the Council of Economic Advisors are all appointed by the president -- more specifically, the chairman is nominated by the president and approved by the Senate. The members are appointed by the president .


Bureau of Labor Statistics Commissioner's Corner: http://beta.bls.gov/labs/blogs/ Twitter Account: https://twitter.com/BLS_gov

mahatmakanejeeves thread - very comprehensive OP each month when the jobs report comes out, as well as additional material he posts to the thread in the following hours. Watch the OP for edits too. And the thread for more material. (In the January 2019 jobs report which came out on February 1, the thread was hosted by UpInArms). https://www.democraticunderground.com/10142257573

Detailed Discussion

2/1/19 -

Nothing much more to add to the above. Usually I'm verbose as most job reports are a mix of good-and-bad when compared to the prior month. But I'm not doing monthly comparisons anymore since the monthly changes in the Household Survey numbers are mostly statistical noise.

Instead, I have been, and will be doing annual updates in early February after the January jobs report comes out (namely detailing the changes over the past 12 months, and since February 1, 2017 which is the beginning of the Trump administration's first full month in office).


##################################################################################
##################################################################################

Beware the tricks of the economic pundits out there, such as right-wingers slamming any gains the economy made under Obama (thanks in large part to Republican obstructionism in Congress, and Republican policies in the many states they control).

NOTE: MANY OF THE BELOW EXAMPLES WERE WRITTEN DURING THE OBAMA ADMINISTRATION WHEN RIGHT-WING GREEDBANGERS WERE DISPARAGING ANY PROGRESS THE ECONOMY WAS MAKING.. Of course they are making the opposite kinds of arguments now that the King of the Magats is president. But the techniques are the same. I have chosen not to spend an extraordinary amount of time re-writing all of the many examples below to illustrate how Magats would try to make the economy look better than it really is.

Tricks of the polemicists include:

(1). Highlighting adverse one-month or other short-term changes in some highly volatile component, and making it seem like it's the story of the whole Obama administration's job record such as, for example, the monthly changes in the civilian labor force, age 16+, seasonally adjusted. Here for example are the monthly changes for 2012 in thousands: (http://data.bls.gov/timeseries/LNS11000000?output_view=net_1mth ):
[font face = "courier new"]
Jan Feb Mar. Apr May June July Aug Sep Oct. Nov Dec
401 498 -15 -246 381 188 -164 -301 349 489 -228 206 (labor force, thousands) [/font]

Needless to say, whenever the president is Democratic, our good friends on the right highlight the drops in the labor force in the months when it drops, and make no mention of the rises when it rises. This is also known as cherry-picking the bad statistic of the month.

As you may know, under Obama, the righties and the media loved to pooh pooh any drop in the unemployment rate when the labor force drops, explaining that the unemployment rate dropped mostly because people gave up looking for work and left the labor force, and so are not counted. But how often have you heard them bring up a rise in the labor force in a month when it rose?

Another example: full-time jobs: [du/10026642259]

Awk! 252,000 full-time jobs were LOST in April! (April 2015, a month where the media was touting the 223,000 gain in payroll employment)

But note this statistic, coming from the Household Survey, is highly volatile from month to month:

Monthly changes in full-time workers (in thousands): http://data.bls.gov/timeseries/LNS12500000
[font face = "courier new"]
` ` ` Jan Feb Mar Apr May Jun July Aug Sep Oct Nov Dec
2014) 410 209 203 396 332 -538 196 310 552 371 -174 427
2015) 777 123 190 -252 (full time workers, thousands) [/font]

One could just as well have said 1,265,000 full-time employees were gained over the last 5 months (253,000/month average). Or 2,314,000 were gained in the last 12 months (193,000/month average). And note that nobody wrote OP's about full-time employment gains when the September report came out (+552,000), or December (+427,000) or January (+777,000)


(2). Cleverly mixing seasonally adjusted data with unadjusted data (without making that clear of course) Or using exclusively seasonally unadjusted data if that paints the picture they want to paint

(2a). Implying that a number is not seasonally adjusted -- for example disparaging a November or December payroll employment report of a good 250,000 increase in payroll employment by saying that's a paltry gain since there should be a lot of Christmas shopping season hiring going on. (Uh, no, like almost all BLS statistics reported in the media, the payroll employment numbers are seasonally adjusted)

Another example - saying a big increase of 0.5% in consumer spending in December is not a big deal, and ought to be way higher since December is after all the big Christmas spending month. (Uh, no, again, the numbers are seasonally adjusted)

(2b). Related -- using NOT seasonally adjusted numbers when that better makes their case, and saying that the unadjusted numbers are "the real numbers" untarnished by bureaucratic "adjustments" and "manipulations"

A great example is comparing not-seasonally-adjusted numbers for December and January, and making an enormous hoo-hah about the decline of the job count in January (when of course much of the extra Christmas season help is laid off, but the polemicists don't mention that explanation).

The below compares the monthly changes of the NOT-seasonally adjusted numbers ( http://data.bls.gov/timeseries/CEU0000000001 ) to the seasonally adjusted numbers ( http://data.bls.gov/timeseries/CES0000000001 ) for the 19 months from January 2014 through July 2015:

Payroll Jobs, Thousands:
[font face = "courier new"]the NOT-seasonally adjusted numbers
` ` ` Jan. Feb. Mar. Apr May June July Aug Sep Oct Nov Dec
2014: -2811 741 957 1163 920 594 -1050 391 687 1081 478 6
2015: -2813 848 779 1139 928 474 -1045

the Seasonally adjusted numbers
` ` ` Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
2014: 166 188 225 330 236 286 249 213 250 221 423 329
2015: 201 266 119 187 260 231 215 [/font]

Of course, most people would see through such an obvious stunt like comparing raw job counts in December and January, so a RW polemicist would rarely try this particular stunt. But I've seen it done for July (a fall off after the hiring of June graduates that most people don't think about). Later: someone at DU actually tried to play this trick in January 2016, sigh.

And they certainly do similarly with other data series where seasonal changes are less understood.


(3). Cleverly mixing statistics from the household survey (CPS) and the establishment survey (CES) (without making that clear of course). The CPS survey of households ( http://www.bls.gov/cps/ ) produces the unemployment rate, the labor force participation rate, the number employed, and innumerable other statistics. The CES ( http://www.bls.gov/ces/ ), a completely separate survey of businesses, produces a number of statistics, most notably the headline payroll employment numbers (widely regarded as a better indicator of job changes than the CPS's Employed number because of the larger sample size among other reasons). Because of statistical noise and volatility, these 2 surveys often come up with seemingly incompatible results. Needless to say, right-wing polemicists mix and match statistics from both surveys to produce nonsense.

(4).Cherry-picking the start and end points of some data series
This is a generalization of item (1.) except that instead of highlighting the latest month of an adverse statistic, they may pick another starting point that is an outlier. For example in October 2013, someone mentioned that the latest U-6 unemployment measure is no better than it was in March 2013, 7 months ealier. True, but March was at a noisy zag low; its clear to see from the graph that there is a downward trend, not a 7-month plateau. U-6: http://data.bls.gov/timeseries/LNS13327709

Here are the 2013 values (the 2012 values are all above 14.4% BTW, it was 15.1% in January 2012)
[font face = "courier new"]
Jan Feb. March Apr. May. June July Aug. Sept Oct.
14.4 14.3 13.8 13.9 13.8 14.2 13.9 13.6 13.6 13.7 (2013, U-6 in percent)
. . . . . . ^-March: the cherry-picked low starting point the RW'er chose[/font]

(The U-6 unemployment rate (sometimes dubbed the underemployment rate) is now (January 2019) 8.1% by the way, down just 0.1 percentage points in the last 12 months (and down 1.2 percentage points in the last 24 months). It is the broadest measure of unemployment that the BLS produces -- it includes part-time workers wanting full-time positions. It also counts as unemployed any jobless person who wants a job and has looked for work at any time in the past 12 months (whereas the headline U-3 unemployment rate counts those who have looked any time in just the last 4 weeks).

It's like global warming when the righties always pick 1998 -- an anomalously hot year because of a strong El Nino -- as their starting point to argue that there has been very little warming since.

That is why seeing the whole data series is so important, and not just accepting the time period and the statistic that a right-wing polemicist dishes out. However, finding the data series number is often quite a challenge, and something that in my experience involves a large bag of tricks. It is my intent to write more about how to find the data series you need. But for now, if there is one trick to mention, this one is the most helpful: http://data.bls.gov/pdq/querytool.jsp?survey=ln

(5). Comparing the current statistics to 2007's statistics, as if 2007 was a normal economy we should get back to - I see this all the time. Yes, today's economic statistics just about across the board suck compared to 2007's. But keep in mind that 2007 was not a normal economy. It was a very sick bubble economy with a very high fever -- people using their houses as ATMs to the tune of hundreds of billions of dollars a year. Anybody could get a mortgage, virtually no questions asked. The belief that housing prices never go down, at least not on a national average scale (thus the theory that a geographically diversified bundle of mortgages was always a safe bet).

The same for comparisons to 2000 -- that too was a very sick economy -- astronomical price/earnings ratios in the stock market, day trading and momentum investing. The belief that Alan Greenspan had mastered the "Goldilocks" economy (not too cool, not too warm) and that, now that we understood how to use the Fed's powers to control the economy, we will never have a recession again. That tech companies with huge negative earnings and no business plan were great investments. That we were all going to the moon, and we were all going to the stars (speaking of the economy and the stock market).

Well, I'm extremely very sorry to have to tell you -- we don't want to get back to the very sick high-fever bubble economies of 2000 or 2007. So quit the whining about how things now are so much worse than back then -- no they aren't when you consider the sickness and unsustainability of those economies back then.

(6). Talking about inflation-adjusted numbers as if they were not, e.g. "wages have been flat (or dropped) for the last 20 (or whatever) years while we all know that prices just keep going up" -- leaving off the word "real" or "inflation-adjusted" qualifier on wages (which takes into account rising prices).

Nominal wages, i.e. raw wage numbers unadjusted for inflation have definitely been rising for years and decades, whereas real wages (meaning adjusted for inflation) have indeed been roughly flat. For example:

(nominal, i.e not inflation adjusted) Average Hourly Earnings Of Production And Nonsupervisory Employees, total private, seasonally adjusted: http://data.bls.gov/timeseries/CES0500000008

(real, i.e. inflation adjusted) http://data.bls.gov/timeseries/CES0500000032 -- to get the long view, set the time period's beginning point from the default value a decade ago to 1964 - the earliest one can set it. The pull-down boxes for setting the time period is near the top, where it sets "Change Output Options". Be sure to check the "include graphs" checkbox, and then click the little dark blue "go" button

(7). Using government statistics and trickery (see above techniques) to make some point, and when you call them on the trickery, and give them the correct information, they tell you they don't trust government statistics! In other words, they are fine with government statistics (or studies that are derived from government statistics, which they all are) if they can twist them to fit their viewpoint, otherwise, they don't trust them!

----------------------------
As for postings by DU members, always check the source of the article they posted, for example one perhaps unintentionally posted a bunch of crap from a right-wing polemicist (Peter Morici) http://www.democraticunderground.com/1251259885#post3 (that's post #3 -- interestingly the poster was PPR'd about 4 months later). Note that sometimes the publication might be an OK mainstream source, but you should still check out the author.


SEE ALSO THE "MYTH:" SECTIONS IN THE PAGES BELOW. For example, the EF-2 page has a lot of myths about the job numbers, such as the myth that they don't count the jobless people that have exhausted their jobless insurance benefits in the unemployment numbers. Just search on the following and include the colon:

MYTH:

Here is a listing of Myth topics as of 1/25/2019

# Myth: "those who have exhausted their unemployment insurance benefits are not counted as unemployed. If they were counted, the official unemployment rate would be much higher" (you often hear this claim from the RepubliCONS when a Democratic president is in the White House, and vice versa when a RepubliCON is in the White House).

# Myth: "But the real unemployment rate is 15% (or 24% or whatever) and it keeps going up":

# Myth: There are 94 million involuntarily unemployed, so the real unemployment rate is about 40%

# Myth: In 1994, during the Clinton administration, they stopped counting the long-term unemployed, or the "long term discouraged worker". If we calculated the unemployment rate now the way we did before 1994, the unemployment rate would be double, triple (or whatever. One claimed that the unemployment rate in January 2015 calculated by the old method would be 23% instead of the officially reported 5.7%).

# Myth: most jobs created during the "so-called recovery" are part-time, especially after Obamacare became law

# Myth: "Those payroll job creation numbers the corporate media reports are just that: jobs created. They don't mention all of the jobs that were lost. To get the true picture, they should report the NET jobs created: jobs created less jobs lost"

# Myth: "The unemployment rate is low because so many people are working two or more jobs, and working 60,70,80 hours a week "


END of "Beware the tricks" lecture
########################################################################


General notes from previous deleted job summaries - to be reorganized and refiled

I'm working on the wages things brought up in earlier DU posts -- for now, See: Real (i.e. inflation-adjusted) average weekly earnings, all employees (total private), 1982-1984 dollars, Seas Adj: http://data.bls.gov/timeseries/CES0500000012
And of production and non-supervisory workers: http://data.bls.gov/timeseries/CES0500000031

Note on statistical noise: As an example: non-farm employment increased by 113,000 in January 2014 in the establishment survey. But according to the household survey, employment that month increased by 638,000. Just goes to show how wild the statistical noise is, and not to get excited one way or another with any one month's particular numbers.

On statistical noise, I found this BLS technical note on sampling error -- http://www.bls.gov/news.release/empsit.tn.htm . Based on what it says, there is a 90% probability that the Establishment Survey's non-farm employment increase is within +/- 120,000 of the stated number. And a 10% chance that it is off by more than 120,000. Again, this is just the sampling error. There are other errors besides sampling error.

Correspondingly, again based on sampling error alone, there is a 50% chance that it is within +/- 49,200. So for example for a reported job gain of 200,000, there is a 50% chance that it is between 150,800 and 249,200, and a 50% chance that it is outside that range based on sampling error alone. Note there are errors other than sampling error that add to the uncertainty {1}

And in the Household Survey, there is a 90% chance that the monthly unemployment change is +/- 300,000 of the stated number (note this is 2.5 times the Establishment Survey's nonfarm employment's sampling error). Also, that there is a 90% chance that the unemployment rate is about +/- 0.2% of the stated number.

The above only covers sampling error. There are also many other sources of error (search the above link for "non-sampling error" )

The individual components that go into these numbers have an even larger sampling error. As explained above, right-wingers love to find the aberrant statistic or two of the month and make it out to be the story of the Obama administration, rather than what it really is -- just one month's number in a very statistically volatile data series.

{1} 90% of the area under the normal curve is between +/- 1.645 standard deviations. 50% of the area under the normal curve is between +/- 0.675 standard deviations. Thus if there is a 90% chance that it is within +/- 120,000, then there is a 50% chance that it is within +/- 49,200 (0.675/1.645 * 120,000 = 49,240 , then round to 49,200 ).


Recent topic updates

# 2/2/18 - All pages (except EF-8 and EF-9) updated -- they were more than a year out of date

# 2/2/19 - ALL PAGES UPDATED TO AT LEAST SOME EXTENT 2/2/19 to mark the 2nd year of the Benedict Donald Trump administration

# EF-0 -
. . . Added 12/10/18: given that there is a 90% chance of the payroll jobs increase is between +/- 120,000 , therefore there is a 50% chance that it is within +/- 47,200

. . . 1/25/19: Added a Myths Table of Contents -- a table of contents of common myths

. . . 5/4/19: Added LFPR and ETPR for several age groups.

. . . 6/27/19: Added Multiple Jobholders as a Percent of Employed

# EF-2 -
. . . 6/12/18: Added Regional and state employment and unemployment http://www.bls.gov/news.release/laus.nr0.htm

. . . 6/27/19: Added Erik Sherman's opinion (a Forbes contributor) why the BLS may be underestimating the number of mutliple job holders.

# EF-3, EF-5 - 4/30/19: Updated GDP and GDP-To-Debt ratios for latest GDP (2019 Q1 First Estimate)

# EF-4 - 8/15/19: Added graph of S&P 500 of the last 6 presidents (Reagan through Trump's first 2 1/2 years). Also improved the formatting of the entire EF-4 "page".

# EF-5 -
. . . Added 4/9/18: Deficit to top $1 trillion per year by 2020, CBO says (2017: $665 B, Projected for 2018: $804 B. The tax cut plus the extra $300 billion in spending over the next 2 years approved in March expected to add $1.6 T to the deficits over the next decade), The Washington Post, 4/9/18

. . . Added 7/1/18: Under current law, national debt held by the public projected to expand from 78% of GDP currently to 152% of GDP by 2048 (30 years from now). By 2048, interest on the debt is projected to nearly quadruple as a percent of GDP (from 1.6% in 2018 to 6.3% of GDP in 2048) and reach the same level as spending on Social Security - CBO 6/26/18 .. ##### On 9/10/18, I found the links were bad on this item and fixed the main one and flagged the spreadsheet one as bad) ### On 10/16/18 I added the Fiscal Year 2018 deficit of $779 Billion below the Imgur table of the last 10 years of outlays and deficits.

. . . 8/31/19: More on why the national debt matters

# EF-9 -
. . . Added 6/2/18: BLS Occupational Employment and Wages Summary - Mean and Median Hourly Wages of Hundreds of Occupations

. . . Added 9/9/18: Minimum Wage

. . . 8/10/19: Added some 5/26/19 wealth inequality statistics

. . . 9/3/19: Added "Historic Asset Boom Passes by Half of Families", Wall Street Journal, 9/1/19

. . . 9/28/19: Added "Income, Poverty and Health Insurance Coverage in the United States: 2018", released 9/10/19.

. . . 9/28/19: Added "US income inequality grows to its highest level in 50 years, Census Bureau, MSN.com, 9/26/19" (it's actually much more than 50 years -- Census Bureau figures only go back to 1967. There's a graph that goes all the way back to 1913).

Romney justifies virtually no job growth at his 3 1/2 year point in Mass.

Probably your crazy uncle is telling you that he's tired of hearing "Oblamer" blame Bush for the poor state of the economy and essentially zero job growth since he took office (since January 31, 2009 thru August 31, 2012, under Obama 261,000 jobs have been lost, although he is in positive territory in private sector jobs).

Your crazy uncle also pooh poohs you when you tell him that in the last 30 months, under Obama 4.6 million private sector jobs and 4.1 million total jobs (actually civilian non-farm payroll jobs), have been created, telling you that you are cherry-picking Obama's best months blah dee blah.

# Payroll Jobs: http://data.bls.gov/timeseries/CES0000000001
# Private Sector Payroll Employment: http://data.bls.gov/timeseries/CES0500000001

Well, Romney in his June 24, 2006 press conference (nearly 3 1/2 years into his governorship of Massachusetts), blamed the economy he inherited for his woeful job record over his entire term, and touted the 50,000 jobs created since the turnaround. Exactly the sort of argument that the righties are criticizing us for making regarding Obama.

Transcript and press-conference video (1:50): http://www.youtube{DOT}com/watch?v=ArRj-dQXX3Y
(replace the {DOT} with . in the above URL. I don't know why it is fighting with me)


TRANSCRIPT:[font color = blue]"You guys are bright enough to look at the numbers. I came in and the jobs had been just falling right off a cliff, I came in and they kept falling for 11 months. And then we turned around and we're coming back and that's progress. And if you are going to suggest to me that somehow the day I got elected, somehow jobs should have immediately turned around, well that would be silly. It takes awhile to get things turned around. We were in a recession, we were losing jobs every month. We've turned around and since the turnaround we've added 50,000 jobs. That's progress. And there will be some people who try and say, 'well Governor, net-net, you've only added a few thousand jobs since you've been in.' Yeah but I helped stop, I didn't do it alone, the economy is a big part of that, the private sector's what drives that -- up and down -- But we were in free fall for three years. And the last year that I happened to be here, and then we turned it around, as a state, private sector, government sector, turned it around. And now we're adding jobs. We wanna keep that going, to the extent we can. We're the, you know, we're one part of that equation, but not the whole equation. A lot of it is outside of our control, it's federal, it's international, it's private sector. But I'm very pleased that over the last a 2, 2 and a half, years we've seen pretty consistent job growth. 50,000 new jobs created, some great companies, we just had, last week, Samsonite announced their headquarters moving here. Companies outside Massachusetts moving in to Massachusetts. That kind of commitment, that kind of decision, says something about what they feel about the future of our state."[/font]

Well, then I wondered, is 50,000 jobs so great for a state the size of Massachusetts? Using July 2011 data, Massachusett's share of the USA population is 6.587 Million / 311.6 Million = 2.1139%. (It would have been better to dig up population numbers more around the 2003-2006 time frame but I doubt that the percentage would be more than slightly different). So on a per-capita basis, 50,000 jobs in Massachusetts is equivalent to 50,000 / 2.1139% = 2.366 Million nationwide jobs.

I'm assuming since the press conference was held in June 24, 2006, that the 50,000 jobs is through the end of May 2006 since they wouldn't have end-of-June numbers in yet.

Well, Obama in a similar point of his presidency -- the end of May 2012, had presided over the creation of 3.744 Million jobs.

So on a per-capita basis since their respective job turnaround points, Obama's job creation record is 3.744 / 2.366 = 1.58 X better (58% better) than Romney's.

And since Romney is "very pleased" with his job creation record in Massachusetts since the turnaround, he should be 1.58 times "very pleased" with Obama's record.


(Note that since Obama took office January 20 (2009) and Romney took office January 2 (2003), I could have moved Obama forward by a month to the end of June. If so, Obama's job creation record through the end of June 2012 is 3.819 Million jobs, and his per-capita record is 3.819 / 2.366 = 1.61 X better (61% better). But I'll settle for the end of May figures. )
Go to Page: 1