Reported: unemployment statistics look terrible
So what happened? An incorrect interpretation of the data – once again.
The positive facts, the negative analyzes and the bad titles
Let’s start with the bad titles …
- The Wall Street newspaper – “The increase in weekly unemployment claims indicates the recovery of fake jobs”
- CNN affairs – “American unemployment claims are rising again for the first time in 4 months”
- Yahoo Finance – “Unemployment Claims: Another 1.416 million Americans filed new unemployment claims last week as the recovery in the job market spits”
- USA today – “‘There are simply no jobs’: the number of Americans applying for unemployment has increased for the first time since March, even if aid is set to decrease”
Now for the bad analysis …
That’s all. Only two numbers. Don’t look back on the trend of several weeks.
Worse still, what happened to everyone who made initial complaints in the previous weeks is not taken into account. The unemployed are still tracked and reported by the “continuous credits” that they must present weekly to continue receiving benefits. The importance of these numbers is that they reveal the dynamics of the unemployed, especially those who have moved on, mainly to new employment.
Finally, the wrong data
The use of seasonally adjusted data is not considered. It is improper because seasonal adjustments are based on normal seasonal occupational shifts and 2020 is wildly abnormal. Therefore, the huge number of unemployed people for coronavirus arrest should not be inflated or deflated by previous seasonal trend adjustments.
Now to the misinterpretations of the latest weekly unemployment report
First, incorrectly applied seasonal adjustments resulted in an increase from a decrease
Start with normal …
In normal times, seasonal adjustments allow for better short-term comparisons, where seasonal factors produce regular cycles. The following chart shows the weekly adjustment factors for initial unemployment benefits, which span the past three years, overlapping from 2020 to July 18.
Now to 2020 …
It is the 2020 unemployment adjustments that are wrong. What we are witnessing are the enormous and abnormal effects caused by the commercial conditions caused by the coronavirus, not by the usual seasonal hiring and dismissal schemes associated with holidays and vacation periods.
Therefore, the real measure of what is happening now is the not– figures on claims adjusted for seasonality. There is no logic to support the inflating or deflating of those anomalous numbers from those normal seasonal factors. (Look at the huge future increases, up to 30%, that come in the normally low unemployment summer weeks. Now imagine the scary headlines that come when seasonally adjusted amounts are reported.)
The following two graphs show the expected seasonal adjustments if the NSA’s initial complaints were 1 million per week (down from an average of 1.4 million in the past five weeks). Note the increases that occur over the next 14 weeks to accommodate the normally low number of requests. Clearly, increasing NSA numbers is wrong, but those recalculated SA numbers will be used by the media as evidence of more difficult times.
Now to the data behind the report on Thursday (July 23) for the week ending July 18 …
Previous weekly (11 July) non-weekly initial requests (NSAs) were 1.51 million. However, this reading was seasonally adjusted (SA) down by -13.6% to 1.31 million.
In the recent weekly report (July 18), the one that produced the negative media reports, the NSA’s initial complaints were 1.37 million, well below the NSA’s 1.51 million the previous week. However, it was adjusted upwards by 3.3% to 1.42 and then compared to the figure 1.31.
Hence, the current weekly drop of -0.14 million has been transformed into a misleading increase of 0.11 million, the fodder for the many negative ratios (even called by some a “peak” of unemployment).
Furthermore, as the predominant seasonal adjustment has been on the rise throughout the coronavirus period, the cumulative number of initial claims has been an increasingly overestimated amount. Here is the graph showing the two cumulative amounts: NSA and SA.
Secondly, attention to initial requests (new unemployed) ignores the strong employment dynamics at work
Initial claims do not mean unemployed forever. Once a person is unemployed, something happens – very often a new job. And this happened regularly even during the coronavirus period. But to get to that dynamic, we need to go beyond the initial claims report and, above all, the cumulative amount of the initial claims.
This means looking at the other often overlooked figure “continues [weekly unemployment] declarations. “Although they appear in the same report as the initial declarations, they are one week late, so they tend to be overlooked. However, they are critical to understanding the unemployment picture.
Start with initial complaint data and continue
This first graph shows initial cumulative NSA complaints (orange line) since mid-March, when the arresting effect of coronavirus took hold. NSA’s continued claims (green bars) show that the current number continues to receive unemployment benefits.
Note the huge difference now between all those who filed initially and those who are still filing. As of 11 July, the initial cumulative complaints were 47.7 million, but the continuous complaints were “only” 16.4 million, with a drop of almost 2/3. Clearly, many people have found work and income elsewhere.
But do those high numbers (1-1 / 2 million initial questions and 16 million continuous questions) indicate that unemployment is still a problem?
Yes, but neither of the two numbers is stagnant. There are clearly dynamics in place when an initial inflow of 1-1 / 2 million does not increase continuous requests. It necessarily means that there is a similar 1-1 / 2 million (about 10%) that leaves the pool of ongoing requests every week. This outflow is the great unmentioned good news that is rarely (ever?) Mentioned.
So here is a chart that clarifies this problem. The green bars are the continuous claims and the orange line shows the cumulative departures of people from the unemployed pool. Note that the ratio is ~ 2: 1, which is the number of previously the unemployed are double the number of unemployed still.
So what is the conclusion from all of the above?
Ignore the weekly message of “Again, __ million new unemployed!” It is both overrated and overly simplistic. To see the real unemployment trend, focus on the numbers ignored by the media:
- Non seasonally adjusted initial credits (NSA)
- The continuing claims of the NSA
- The outflows of former unemployed
And here’s another thing, not discussed above …
Study the monthly “Employment situation“Report from the Bureau of Labor Statistics of the United States. A complete data collection process is conducted every month, then analyzed and reported in two phases: preliminary results in the following month and final results in the following month. (August 7 is when preliminary July Statistics for 2020 and for the last June 2020 are reported).
BLS provides unemployment and unemployment by age, education, gender, ethnicity, etc. Unemployment numbers are total, regardless of whether the unemployed receive benefits or not. (The graph below shows the difference in the numbers.)
The examination of the report reveals long-standing facts, current trends and a reality that is lost with simplified communication to a number (for example, the unemployment rate of 11.1%).
The bottom line: we have to conduct our analysis of unemployment
For an accurate view of the unemployment picture, we need to ignore the short-sighted view of the media about a week’s change in seasonally adjusted initial unemployment claims.
Why do we have to do it alone? Because today’s news works according to the 15-minute rule. Publish a news report more than 15 minutes after the event (for example, releasing weekly complaint data) and the number of viewers decreases significantly. Obviously, this leaves no time for a thoughtful analysis. It is also the reason why headlines look so similar nowadays.