Sunday, August 26, 2012

Burden of Debt, Third Post (and last for now...)

Yesterday I stopped with a promise to show a more general, or maybe more realistic, version of the simple diagram I was using to illustrate Nick Rowe’s point about the burden we can place on future generations by borrowing today even if we never place a burden on future time periods.  I’ll do that here, but first I want to make quick point: this is not just an exercise.  The points do have relevance to current issues even if the diagrams looked simple.  For example, the last point I made yesterday was that economic growth can make the issue of inter-generational equity a lot muddier, and can make a current plan that imposes an extra tax on future generations look inter-generationally equitable.   This is the exact issue that Dean Baker was raising when he said this:

However, the change in the projected growth of health care costs also has another much more important implication that went altogether unnoticed. It means that workers in the future will be considerably wealthier than we had previously believed. In other words, if healthcare reform will effectively contain cost growth without jeopardizing quality, then our children and grandchildren will be far wealthier than in a world without health care reform.

The 2010 projections show the average worker's wage will be 47.8 percent higher in 2040 than it is today. This is after adjusting for inflation, so the projections show that workers' purchasing power in 2040 will be 47.8 percent greater than it is now. The new projected annual wage for 2040 is 6.3 percent higher than figure projected for last year.

To understand the importance of this change in wage growth projections, suppose we told our children and grandchildren that the payroll tax would have to be raised by 3.0 percentage points to support Social Security (an extraordinarily large increase). They would have more money in their pockets with the tax increase under the current projections, than with no tax increase and the wage growth projected in the 2009 report.”  

So if we can make our simple diagram more realistic, and provide some more accurate data to fill it, we might be able to see something about our current decisions that matters.  I don’t plan to do all of that here, because it sounds like a career rather than a blog post.  But I’ll take one step in that direction.   Actually, this is an area where I haven’t done a lot of reading, so I’m sure there are better versions somewhere on the web of the graphs I will show here, but I haven’t been able to find them in what I admit was a fairly brief search.  Maybe someone who reads this can point me in the right direction.

An initial word of caution is in order: I grabbed the easiest real world data I could find to do this.  That turned out to be the median household income by age group, which I got from the Census Bureau in the Commerce Department historical data on incomes here.  I looked at table H-10, all races, to grab some numbers to fill a graph that showed the median of income by age of head of household by year, that is, by what I was calling an accounting period in my last post.  That may not be the best data to use for this---I mean, why the median instead of the mean, for example? And since the data I had showed things like “15 to 24, 25 to 34” and so on, I assigned the number for that category to the middle of that range---meaning I said that the number for “15 to 24” was what a 20 year old head of household made.  So that gave me data for ages 20, 30, 40 and so on. Finally, in order to make the data fill the graph I had to do a lot of interpolation to give me a data point for each year of life, which I did in the easiest and probably least justifiable possible way, which was straight lines between the points. Because of this, and knowing that once you put something on the web it can go anywhere and be used for anything, I’ve embedded nasty warnings at the tops of these graphs that they are not to be used for argument in favor of against any specific policy agenda.  A graph like this can be used, but first it has to be filled with trustworthy data that is appropriate to the task.
Here’s the result (below).  This is very like the simple two-age diagram from yesterday’s post, except I have ages from 20 to 70, and years 1967 to 2010.

I think this kind of diagram can be useful---I’m tempted to use the phrase “no matter how you slice it”, because that’s what I want to do.  The first slice direction is already there in the diagram: that’s the slice-by-year or slice-by-accounting-period that is shown as the various color slices in the picture.  The purple slice in the front is the data for distribution of income in 1967.  It has the shape you might expect: young heads of household have modest incomes, but that grows through life until they get to be about 50 when some of them begin to retire.  After that it falls off fairly rapidly, and the lowest lifetime income occurs at (or after) age 70.  In some ways, that might even be optimal: it’s in the middle of life, from 30 to 55, that many people have the heaviest responsibilities for supporting their children, or for helping their parents and their ne’er-do-well rapscallion siblings. 

A second slice direction would be to slice along the “year” axis.  You can see one of those slices by looking along the edge of the picture, at age 20.  But here’s a version that shows several such slices:

Interesting.  People who were seventy in 2010 were doing a lot better than people who were 70 in 1967.  The elderly have done fairly well over the last 40 years.  The rise in their incomes was not impacted much by the dismal economy of the late seventies and early eighties, but it was also not accelerated by the Volcker recovery from the Volcker recession through the mid-eighties.  Whether you think this is a good or a bad thing is a matter of personal vision. On the other hand, the thirty year olds have been buffeted by every recession, early eighties, early nineties, 2000, and the most recent one starting in 2007.  People who were 30 in 2010 were better off than those who were 30 in 1967, but not by much.   And so on.

Now let’s look at the cohort slices, the 45 degree slices I sketched on top of the diagram in colored lines.  Here’s what those look like when you slice in that direction:

This is a picture of the trip through life of particular groups, particular cohorts.  People who were 20 in 1967 are the blue line on the top.  The fact that they are on top, and that the lines nest underneath each other in age order, is evidence that things do get better; that in spite of everything each cohort so far is better off than the cohorts before them.  The blue light line at the bottom is the group that was 60 in 1967.  The teal line above that is the group that was 50 in 1967.  And so on.  The right side edge of the graph is the income---or their rough estimate within my massaged data set---of each of these cohorts at age 63, and each of the lines start in the year 1967, so that the ages of the cohorts depicted in these lines will line up.   One thing this means is that events in time, like the early-eighties recession, occur at different points along the different lines.  That event, for example, is the downward kink in the dark blue line at roughly age 32, and in the yellow line at roughly age 42, and so on.  You can see that the cohorts that were in their thirties or forties when that recession started had a reasonable recovery---but those that were in their fifties or older (the purple line at roughly age 52) basically never recovered from it.   I think that may be true in every recession.

One other thing that’s immediately evident from this picture is that life is not a smooth journey: all that bumpiness, the ups and downs of the lines, shows the turbulence of the times these cohorts lived through.  And this was a very easy time with limited economic turbulence!  This graph covers the time period that economists have been calling the “great moderation”!  I would guess that the life-lines of cohorts that lived through the world wars, or the great depressions, or the dust bowl, were vastly more turbulent than these.

You may have noticed that I haven’t talked about slicing the NICE graph along the vertical axis, that is, along income slices.  That’s partly because I don’t know what to do with that graph.  Economists use contour maps of mountains a lot (ask your local economist about isoutility curves or isoquants or iso anything else: they are all contour plots of mountains of utility or quantity or something.)  But I’m not sure it shows anything new or enables anything we haven’t already seen.  But it’s worth plunking a contour map down here just to be complete.  So here it is:

There is one thing that’s easier to see in this view than in any of the others.  If you were 20 years old in 1971, your cohort went right through a sweet spot in this graph, the highest median household income shown.  If you were 20 in 1988, you seem to be moving through a mountain pass: your cohort seems to be skirting that peak of incomes. 

Recessions hurt.   But they don’t hurt everyone equally.
So I’m back to where I started, in my opinions, in a way.  Does debt “burden” future generations?  It depends.  But failing to create sufficient stimulus is burdening our current generations, all of them that are around right now.  We have a lot of idle resources, and record low interest rates---actually negative real rates on Treasury bonds.  We can incur debt now and invest in infrastructure and education, and the end result will help us all get up off the floor of this mountain pass right now, and will lay the foundation of prosperity in the future. 

Krugman wrote a book with this title: End This Depression Now!   It’s a good book.  I recommend it.  We can, and in this case, under these conditions, that means we should.

No comments:

Post a Comment