Relevant and even prescient commentary on news, politics and the economy.

Demographics and real wage growth

When discussing real income growth one of the points often made by conservatives and libertarians is that the average data on real income is irrelevantly because it does not measure the same people over time. There is some validity to this point. For example, an individual captured in the snapshot of income data in 1970 who is only 20 years old will be 55 in a snapshot of income in 2005. Because income tends to increase over time they make the argument that the average income data understates the income gains people actually experience.

The type of data behind this argument is this data on median earnings of full time employees.

(% of median) (%of 15-24)
total 100
15-24 66.0
25-34 99.8 151.2
35-44 115.6 175.2
45-54 121.6 184.3
55-65 112.7 170.8
over 65 84.2 127.5

It shows, for example that a 50 year old’s earnings are 121.6% of the average and/or 184.3% or a 20 year old’s earnings.

Of course if you have a normal pyramid shaped age structure the snapshot of average income in 1970, 1980….2000 will include the same share of 20,30…60 year olds at each time so the argument advanced that the people change is invalid. But over the last 30-40 years the US has experienced a very unusual demographic shift in the labor force because of the baby boomers maturing. As the chart shows the composition of employment has changed drastically.

But the combination of a changing income age structure and income increasing as the labor force
ages generates a significant impact on average wage growth. If you take the income structure shown in the table and apply it to the changing demographic it generates significant results.
In the 1970s because of the baby boomers entering the labor force the age of employees and the real median weekly earnings of the US employees would have actually have fallen some 2%.
From the bottom in 1980 these factors alone should have generated about a five percentage point increase in real medium weekly earnings–from 98 to 103.3 in the chart.

The actual increase in real weekly median earnings from 1980 to 2007 was about eight percentage points — from 98 to 106.2 in the chart.

So when you actually look at the data behind the claim that average real income comparisons over time are invalid you get some very interesting results. It shows that 64% of the actual increase in real income from 1980 to 2007 was due to the maturing of the labor force rather then other general economic factors such as productivity growth. So rather than productivity, etc. generating an eight percentage point real income increase from 1980 to 2007 it was only three percentage points.

Of course this is just the opposite point the conservative and libertarians intent to make, but
why should we try to confuse them with the facts.

Comments (0) | |


Here are some charts to demonstrate what I was talking about. the recent drop in US bond yields has not been accompanied by a similar drop in European rates.
Consequently the spread between US and foreign bonds has narrowed and it no longer compensates foreigners for the currency risk they take when investing in the US.
The consequences are a weak dollar. Note the opposite happened when Reagan first
created structural deficits and the dollar had to weaken enough to create a large enough
trade deficit to facilitate the capital inflow. At that point we had crowding out but it worked through the dollar to crowd out the manufacturing sector rather then the interest sensitive sectors.

It is also working against the Euro as you see the it strengthen when rates spreads widen.

When the US made itself dependent on foreign capital inflows the world got a lot more complex and it was no longer possible to analysis monetary or fiscal policy as if the US were a closed economy. I worry that we are now starting to see the bear case that I have worried about for years when international capital flows prevent the Fed from easing when domestic considerations call for it. I’m not making a forecast, I’m just laying out factors that make the analysis much more complex and that need to be brought into the analysis.

Maybe it is what we need to revive the manufacturing sector, and will be what drives the economy over the next decade. Save-the-rustbelt would love this. But, it creates inflation and real income complications for the rest of us.

I probably is what will happen since Don Boudreaux at the Cafe Hayek blog just got one of his letters to the FT published where he praised the trade deficit. I just base my analysis on the thesis that he is usually wrong.

I realize these charts are simple and the real world is more complex involving covered interest rate difference, etc, but they are adequate to make the point.

Tags: , , Comments (0) | |


Those who do not study history are doomed to repeat it.

Just to put things in perspective, the inflation rate was 4.4%, the same as it is right now, when Nixon imposed price controls in 1971.

P.S. He knew the inflation rate was slowing. But he imposed price controls as part of his package to devalue the dollar because he expected the weak dollar to be inflationary.

Tags: Comments (0) | |


Cool it everybody. If you look at the historical data you will see that it is a very normal cyclical development for the participation rate to drop in the early stages of a recession. There is nothing funny oR out of line with the data.

Comments (0) | |

Employment report

The February index of aggregate hours worked was 107.3 as compared to a 4th quarter average of 107.7. That implies that first quarter hours worked are falling at about a 1.5% annual rate. As a first approximation that implies that first quarter real GDP growth will be about the same. Of course this depends on productivity growth. if you assume 1% to 2% productivity growth you are back to around a zero growth rate for real GDP. It looks like a recession, but it could still just be stagnation.

Average weekly earnings rose at a 3.3% annual rate. This implies that real earnings are still falling so there is little hope for a turn around in consumer spending even though February same store sales growth appeared to improve and February auto sales were 15.38 M (SAAR) as compared to 15.33 M in January.

Comments (0) | |

Capital Gains and Dividend Taxes

We keep hearing from the conservatives-libertarian that we need lower taxes on capital gains and dividends to encourage capital spending or investments. But when I look at the record of the impact of lower taxes on capital on investment I am hard pressed to see the benefits such lower taxes are suppose to deliver.

The peak was under Carter just before Reagan cut the capital gains and dividend taxes.
Clinton raises them and Bush cut them again. it sure looks like the cut in taxes lead to lower investments, not the higher investment the tax cuts were suppose to generate.

Yes, the data is nominal data. But tax rates work on decisions made on nominal dollars so looking at their impact on nominal dollars is the correct thing to do.

Tags: , Comments (0) | |

Health Care Spending

The other point from my earlier chart on health care spending worth noting is that in the PCE data health care accounts for around 20% of consumer spending. In contrast, medical care only has a weight of 6.2% in the CPI. Since health care prices generally rise some two to three percentage points faster then overall inflation this implies that the CPI significantly understates reported inflation. If the weight of medicine in the CPI were more like the weight in spending, the reported inflation rate would have been up to a full percentage point higher every year.

Interestingly, when I bring this point up to all the people who argue that the CPI overstates inflation they generally try to change the subject.

Tags: Comments (0) | |

Health Care Spending

I was looking at some data yesterday and encountered a big surprise. I suppose like almost everyone I was under the impression that health care spending was growing much faster then the economy and absorbing an ever growing share of resources. This theme dominates much discussion of health care issues, particularly projections of the federal budget. So I was very surprised to look at the data and find that this while this assumption was very valid from 1950 to the early 1990s, there was a very sharp break in trend in the early 1990s and health care spending relative to the rest of the economy slowed very sharply — see the chart.

UPDATE: This data is from the GDP accounts and is PCE on health care services. I went back to the data and found that drugs were not included. Drugs were reported under nondurables. So
here is what the data looks like when I add drugs into the data. It does not change my problem much.

This chart implies that despite the aging of the baby boomer generation that some of our fundamental thinking about the course of health care spending may be inaccurate. This data raises some very important questions in my mind that maybe some readers of this blog that are more knowledgeable may be able to answer. Part of the answer appears to be that the rate of price increases in health care has slowed sharply.

Bruce Webb or others that have looked at this issue, can you help me.

update: The other interesting point is that the creation of Medicare in the 1960s did not appear to impact the trend that had already been in place since 1950, long before Medicare was created.

Comments (0) | |