There was "good news" in the U.S. in October 2019 (see article). The U.S. added 128,000 jobs to the economy. Unemployment is down.
But, wait, what ARE those 128,000 jobs? According to CNBC, "The
Labor Department explained in a press release that much of the
upswing in leisure and hospitality came from a surge in hiring at
food and beverage establishments, which alone added more than 45,000
jobs for the sector’s best month since January" The article
further indicated that 60,000 were from the leisure and hospitality
sector while health care and social assistance added 34,200 jobs. Is
there a pattern here? Ah -- 22,000 jobs were created for professional
and business services which is defined as management consultant
positions, computer system design, and architectural and engineering
services.
So, we have 105,000 jobs added in fairly low-wage areas, 22,000 added in relatively high-wage areas, 36,000 lost in medium-wage manufacturing, and a few dozen thousand added in low medium-wage areas. Is this a success? Or is it a failure?
The point is that the number of jobs added is not very relevant unless one also knows the median wage paid for those jobs. In the U.S., and in most of the more industrialized areas of the world, "income inequality" continues to accelerate. Numbers such as the "number of jobs created" are meant to encourage and say things are great. Maybe they are -- but the analysis of the October 2019 job results don't tend to lend themselves to that conclusion. Lots of low-paying jobs and many fewer higher-paying jobs do not start to spread the economy out to more and more people.
"The average wage for people in the U.S. has gone up!" Sounds great. Wait a minute. If I have 10 people working and five earn $10, 3 earn $15, 1 earns $20, and 1 earns $100 -- we have an average wage of $21.50. What happens if that person earning $100 now earns $200? Wow. We now have an average wage of $31.50. Doesn't that sound great! probably not for the lower-earning 9 people.
So average doesn't tell us much. How about median or mode? Median is the value given when, sorted in order, the number of instances of a value below a point is the same as that above that point. In our example above, with an even number of instances (10), the median is the average of the values at position 5 and position 6 -- so, for the above example, it would be $12.50. Note that, unlike for average, if the person in position 10 (highest earner) doubles their income, the median does not change.
How about mode? Mode is the "most popular" instance. Since, in the above example, five people earn $10, $10 is the mode. Once again, when the most highly paid person gets paid double, the mode remains the same.
There is a quote popularized by Mark Twain which I have always liked: "There are three kinds of lies: lies, damned lies, and statistics". Although this does call out the need to both understand, and interpret correctly, statistics -- it is not really fair to statisticians. But statistics are a politician's best friend largely because most people do NOT know what statistics mean or how they are calculated.
"The average wage has doubled!" Sounds good, right? Not necessarily. "500,000 jobs have been produced in the past month!" Wow, things are completely rosy. Once again, not necessarily. If 400,000 jobs were lost then it is actually terrible. Usually, what is reported is NET gain so if 500,000 jobs were added and 400,000 were lost it would be reported as 100,000 jobs added. Sound good? Well, with that great of a turnover of labor, the chances are not wonderful that those 400,000 (actually, probably not the same people) who lost and got a job improved their wages and benefits.
"$70 million dollars of welfare fraud found". Horrible, right? Well, yes, it should certainly be reduced if possible but total welfare funding is about $67 billion. So, this atrocious fraud is actually part of 1/1000 of all of the welfare funding -- or 999/1000 are NOT committing fraud. In most businesses, this would be a resounding success. In Florida, politicians stamped the ground and screamed about all of the welfare fraud -- and spent MORE money investigating the fraud than the net amount of the fraud. A net taxpayer loss.
Let's take the above $67 billion in welfare funding mentioned, once again. A lot of money isn't it? It certainly is and I would not mind getting my $200 share of that money. But, within the context of overall spending, the $67 billion is insignificant in comparison to the overall federal budget -- about 0.5% or 1 part out of 2000 [note that it is about 21% of the federal social support budget but that is also not a high percentage of the overall federal budget]. The $67 billion is not significant overall but politicians blow it up to seem significant so that people will ignore the other 99.5%.
Another side track of statistics from the politician's mouth is rankings. We are number one! OK. We probably are -- but number one in ranking for what characteristic? Hopefully, the people of most countries think that their country is the best -- a subjective ranking that has multiple truths based on who is saying it and where they are saying it. But, when we get to rankings based on quantitative (countable) characteristics, the record isn't so great. Yes, if a specific formula is applied to the countable numbers, there will be a most highly ranked (# 1) and a most lowly ranked (# xxx). But not many countries have multiple "# 1 rankings" for their country. Given a large enough population and land mass, most countries will have A (one or more) "#1 ranking". But, even in this case is being first in numbers of prisoners equivalent to being first in the number of books per household? So, relative rankings are fairly useless -- dependent on many different factorsl
Numbers do mean something -- but not always what they are implied to mean.
So, we have 105,000 jobs added in fairly low-wage areas, 22,000 added in relatively high-wage areas, 36,000 lost in medium-wage manufacturing, and a few dozen thousand added in low medium-wage areas. Is this a success? Or is it a failure?
The point is that the number of jobs added is not very relevant unless one also knows the median wage paid for those jobs. In the U.S., and in most of the more industrialized areas of the world, "income inequality" continues to accelerate. Numbers such as the "number of jobs created" are meant to encourage and say things are great. Maybe they are -- but the analysis of the October 2019 job results don't tend to lend themselves to that conclusion. Lots of low-paying jobs and many fewer higher-paying jobs do not start to spread the economy out to more and more people.
"The average wage for people in the U.S. has gone up!" Sounds great. Wait a minute. If I have 10 people working and five earn $10, 3 earn $15, 1 earns $20, and 1 earns $100 -- we have an average wage of $21.50. What happens if that person earning $100 now earns $200? Wow. We now have an average wage of $31.50. Doesn't that sound great! probably not for the lower-earning 9 people.
So average doesn't tell us much. How about median or mode? Median is the value given when, sorted in order, the number of instances of a value below a point is the same as that above that point. In our example above, with an even number of instances (10), the median is the average of the values at position 5 and position 6 -- so, for the above example, it would be $12.50. Note that, unlike for average, if the person in position 10 (highest earner) doubles their income, the median does not change.
How about mode? Mode is the "most popular" instance. Since, in the above example, five people earn $10, $10 is the mode. Once again, when the most highly paid person gets paid double, the mode remains the same.
There is a quote popularized by Mark Twain which I have always liked: "There are three kinds of lies: lies, damned lies, and statistics". Although this does call out the need to both understand, and interpret correctly, statistics -- it is not really fair to statisticians. But statistics are a politician's best friend largely because most people do NOT know what statistics mean or how they are calculated.
"The average wage has doubled!" Sounds good, right? Not necessarily. "500,000 jobs have been produced in the past month!" Wow, things are completely rosy. Once again, not necessarily. If 400,000 jobs were lost then it is actually terrible. Usually, what is reported is NET gain so if 500,000 jobs were added and 400,000 were lost it would be reported as 100,000 jobs added. Sound good? Well, with that great of a turnover of labor, the chances are not wonderful that those 400,000 (actually, probably not the same people) who lost and got a job improved their wages and benefits.
"$70 million dollars of welfare fraud found". Horrible, right? Well, yes, it should certainly be reduced if possible but total welfare funding is about $67 billion. So, this atrocious fraud is actually part of 1/1000 of all of the welfare funding -- or 999/1000 are NOT committing fraud. In most businesses, this would be a resounding success. In Florida, politicians stamped the ground and screamed about all of the welfare fraud -- and spent MORE money investigating the fraud than the net amount of the fraud. A net taxpayer loss.
Let's take the above $67 billion in welfare funding mentioned, once again. A lot of money isn't it? It certainly is and I would not mind getting my $200 share of that money. But, within the context of overall spending, the $67 billion is insignificant in comparison to the overall federal budget -- about 0.5% or 1 part out of 2000 [note that it is about 21% of the federal social support budget but that is also not a high percentage of the overall federal budget]. The $67 billion is not significant overall but politicians blow it up to seem significant so that people will ignore the other 99.5%.
Another side track of statistics from the politician's mouth is rankings. We are number one! OK. We probably are -- but number one in ranking for what characteristic? Hopefully, the people of most countries think that their country is the best -- a subjective ranking that has multiple truths based on who is saying it and where they are saying it. But, when we get to rankings based on quantitative (countable) characteristics, the record isn't so great. Yes, if a specific formula is applied to the countable numbers, there will be a most highly ranked (# 1) and a most lowly ranked (# xxx). But not many countries have multiple "# 1 rankings" for their country. Given a large enough population and land mass, most countries will have A (one or more) "#1 ranking". But, even in this case is being first in numbers of prisoners equivalent to being first in the number of books per household? So, relative rankings are fairly useless -- dependent on many different factorsl
Numbers do mean something -- but not always what they are implied to mean.
No comments:
Post a Comment