Tuesday, June 16, 2015

Four percent?

Timothy Noah from Politico called yesterday to ask if I thought four percent growth for a decade is possible. Story here. In particular he asked me if I agreed with other economists, later identified in the story, who commented that it has never happened in the US so presumably it is impossible.

This prompts me to look up the facts, presented in the charts at left. The top graph is annual GDP growth. The bottom graph gives decade averages. Data here. The red lines mark the 4% growth point. Notice the sad disappearance of growth in the 2000s.

Judge for yourself how far out of historical norms a goal of 4% growth is.

By my eye, avoiding a recession and returning to pre-2000 norms gets you pretty close.  A strong pro-growth policy tilt, cleaning up the obvious tax, legal, and regulatory constraints drowning our Republic of Paperwork (HT Mark Steyn) only needs to add less than a percent on top of that. 4% might be too low a target!

Note the question asks about real GDP not per capita. Adding capitas counts. If you want total GDP to grow, regularizing the 11 million people who are here and letting people who want to come and work and pay taxes counts toward the number. You may argue with the wisdom of that policy, but the point here is about numbers.


This is not a serious answer to the question whether 4% growth is possible. The serious answer looks hard at demographics and productivity, estimates how far below the free-market nirvana level of GDP we are, and estimates how much faster free-market nirvana GDP could grow.  If you think that sand in the gears or inadequate infrastructure or not enough stimulus means we're 20 percent below potential, and potential can grow 2% per year, then 4% growth for a decade follows.

What happened in the past is largely irrelevant, since the US has never experienced free-market nirvana. If you were to look in 1990 at historical Chinese GDP plots to assess whether it is possible for China to grow as it has for the last 25 years, you'd say it's impossible. Conversely, if you were to look at postwar US data you'd say our lost decade of 2% growth can never happen.

But, having asked the question whether four percent is outside of all US historical experience, at least it's interesting to know the answer.

Update. A few commenters asked about my 10 year average plot. I've updated it to include two ways of calculating the 10 year growth rate. The solid line is the ten year percentage growth rate divided by ten. The dashed line is the ten year average of one year growth rates. You can also haggle about compounding,  logs vs levels, which price index to use, long-term inflation measurement (that can add as much as 1%), whether some components of GDP should be excluded and so on. Calculated either way, I conclude that 10 years of 4% growth is not an outlandish impossibility. After all, the policy choices being advocated for 4% growth go a good deal beyond rewinding the clock to exactly the policies and laws of some bygone era. But make up your own minds.

Complete disclosure:

dgdp = 100*(gdp(2:end)./gdp(1:end-1)-1);
dgdp10 = 10*(gdp(1+10:end)./gdp(1:end-10)-1);
dgdp10a = filter(ones(10,1)/10,1,dgdp);





37 comments:

  1. "There is a whole Bush Instiute 'four percent' project, curiously unmentioned here."

    Hmmm. In the Politico piece Noah mentions, interviews and quotes the Bush Institute over four paragraphs (starting "Bush didn’t say precisely how long he expects the economy to grow at 4 percent. But the Dallas-based George W. Bush Institute....").

    ReplyDelete
    Replies
    1. Thanks for pointing out the error. Long story.

      Delete
    2. Please read the book The 4% Solution. It is all in there.
      Jeff Cunningham

      Delete
  2. > This is not a serious answer to the question whether 4% growth is possible. The serious answer looks hard at demographics and productivity,

    So ... the serious answer would be no? (with no dividend from women entering workforce, aging baby boomers leading to flattish working-age population, questionable productivity?)

    ReplyDelete
  3. Nice Graph. Thanks. Yes, 4% not likely. But there is room as labor participation rates are still relatively low - even after accounting for "baby boomers, etc." "Growth-friendly" policies, as Dr. Cochrane ponts out, would help both job growth and productivity growth. Under such circumstances, with perhaps a boost from immigration as well, higher than "expected" growth would be quite possible.
    .

    ReplyDelete
  4. Do what America did to Japan.

    Drop a few atomic bombs, wipe out 80 per cent of industrial production, have a Marshall Plan for recovery, and a baby boom following the demobilisation of soldiers, you've got it. It's 1955 again.

    Seriously though, it would be good if growth economists put the models away for a while and did some serious historical study on how some economies went through exceptional periods of growth at certain times. Piketty and Maddison I think are historically well-informed and good guides to the long term trends of capitalism.

    ReplyDelete
  5. I don't like you averaging method, I referencecyour time series manual for every problem so have, I was expecting to be wowed

    ReplyDelete
    Replies
    1. Wow I need to learn to type on phones:
      "Reference your" "problem I have"

      Delete
    2. 1 / N sum i = 1 to N. What's not to like?

      Delete
    3. Maybe averaging method isn't the right term...it looks like you are measuring growth by dividing by ending date GDP rather than starting date GDP...don't we usually do it the other way around...both skew the graphs a little so we move onto other smoothers...

      Delete
    4. Never mind my excel had the wrong formula

      Delete
  6. No wait...I retract my retraction...

    ReplyDelete
    Replies
    1. I updated it using both ways and providing the formula.

      Delete
    2. Nope I can't match your graph...I'm chalking it off to graphing software rounding error, I tried logs, percent changes and some more exotic looking conspiracies from the blogosphere...none of mine seem to hit 4percent where yours do circa 2000....

      Delete
    3. Now I'm matching...I missed a row somehow...

      Delete
  7. You can't ignore the shocking correlation between the rate of growth and the political party in control. During years when Democrats have controlled both the white house and the legislature, we have AVERAGED 5.2% GDP growth. During the (admittedly few) years that Republicans have had control of both those branches, they have never hit 5% at least since 1930- http://politicsthatwork.com/graphs/gdp-growth-by-party

    During years when control is split, we've averaged 2.3%, which is just about where we are now.

    Voting Democrats in definitely isn't the whole game, but it seems to be a big part of it.

    ReplyDelete
    Replies
    1. Yes, because the Democratic and Republican parties have not changed at all in the last 85 years. You could just as easily look at control on congress since about 1980 and find the opposite correlation. Not that this has anything to do with economics anyway.

      Regarding the actual post, after the recession ended weren't forecasts from the CEA and the Fed more or less projecting gdp to return to it's previous trend within a few years? Maybe 4% is a bit ambitious, but why is it suddenly so clear that 2% growth has to be the new normal?

      Delete
  8. 4% would be an unexpected and nice surprise! Either way, it can trend higher if we make some of the changes you discuss as well as others. 3,3.5, or 4 all sounds good from here.

    ReplyDelete
  9. Dr. Cochrane,

    The periods when 4% occurred were mostly before 1980 and entirely before 2000. In the 1950's the manufacturing sector yielded between 40 to 50 percent of all corporate profits and employed about 20 to 30 percent of all The workforce. Finance, insurance, and real estates (FIRE) has never employed more than 5 percent but today yields just a wee bit less than 30 of all corporate profits.[1] Different studies vary on exactly how few people the financial sector employs and how much profit they produce but they all produce similar results [2]. No other sector of the United States economy produces so much profit on a percentage basis but employees so few [3].

    Periods of 4% growth all occurred when the economy of the United States was dominated by manufacturing. As manufacturing declined, the frequency of periods of 4% growth declined. In a heavily financialized economy, one with declining manufacturing, 4% growth is well nigh impossible.



    [1] http://bit.ly/1zW5vPW

    [2] http://bit.ly/1zW5xYe

    [3] http://reut.rs/WWnQsR

    ReplyDelete
  10. The 1960s was our most productive decade of the 20th Century and since, by any measure of GDP you wish to make -- I used per capita (real) GDP -- and there isn't another decade that comes close. Why?

    (1) The composition of the economy was more balanced between Govt spending (less than 35% for Fed, State, and local), the consumer sector (about 40%), and the industrial/commercial sector (about 25%). (2) The combination of these sector gave a higher payback, many of our industries was home-based, than what we have today. (3) The compostion of the economy, with many industries home-based, gave a greater turnover of dollars. (4) Federal Government policy, to inlcude the Fed, could more effectively pump-prime the economy with fiscal and monetary policy -- these tools are very limited today and blunt instruments for public policy.

    ReplyDelete
    Replies
    1. My explanation for the 1960s would be: catch up investment in technologies developed since 1930, baby boomers, women's lib; transistors and other technological developments (e.g.,jet engines) that had been funded by the government in the 1950s.

      Delete
    2. (1) If catch up technologies was a factor it should influnce the 1950s like the 1960s -- during the '60s there was a marked increase in expenditures for Plant & Equipment over the 1950s, when we were coasting on Europe still making a come-back from the effects of WWII. (2) The entrance of the first wave of baby boomers would be a factor; but, this should have had a greater effect on the 1980s -- the '60s are 30% to 40% more productive than the '80s depending upon what you use as a base year (the '60s or '80s). (3) Women in the workforce in the '60s would be only slightly greater than the '50s. It was the '70s and since when this became a factor. (4) "transistors and other tech developments" -- this is not only peculiar to the '60s; why not the '80s, '90s?

      The composition of the '60s economy along with the productivity increases allowed for a synergy; e.g., advances with NASA funding/discoveries had eventual spillover effects in other fields like medicine and medical technology. That composition allowed for a more effective implementation of fiscal and monetary policies. And LEADERSHIP shouldn't be overlooked -- JFK, Martin at the Fed, and within the private sector and university system (which was ramped up after Sputnik).

      I would also note that as we went through the '60s, our competiveness declined, especially with continental Europe.

      Delete
  11. Since there has been a lot discussion here lately about writing concisely I took a swipe at rewriting the Politico article, including a new title.

    Economists Weigh in on Whether a Hypothetical Phenomenon that Has Never Occurred Can Occur

    Q: Is it possible to have 4% GDP growth for a decade?

    A: Of course it's possible. Nobody has proven it's impossible. It's just highly unlikely.

    Q: What about the economists who say it's impossible because it's never happened before?

    A: Only idiots and small children think like that.

    ReplyDelete
  12. Interesting post.
    If the US entered a long and large conventional war, would GDP expand by (more than) 4% a year?

    Probably.

    ReplyDelete
    Replies
    1. Yes. I should have added that the '60s were actually second to only the '40s; which was around 4% annually.

      Delete
  13. The difference between GDP and GDP per capita is significant because the birth rate has fallen so far.

    We have had prior periods of high growth in total GDP because of the baby boomers entering the work force, women entering the workforce and the technological leaps of the past. The microprocessor was invented in 1970 - we are still coasting on that one particular leap forward.

    You want higher growth: do justice reform so that fewer people are in prison and more people are in the work force; and pour money into research so that we can move the technological frontier.

    ReplyDelete
  14. American GDP is under reported as a result of the widespread use of tax schemes that artificially report in tax shelters profits which in truth are generated in the United States.

    ReplyDelete
  15. BTW--the real GDP did in fact grow by 20% from 1976 through 1979. The Jimmy Carter years, when the Fed printed a lot of money. History is ugly.

    ReplyDelete
    Replies
    1. Using Real GDP, 2009 $s, 1976 - $5.675 trillion; 1979 - $6.466 trillion -- 14% over 4 years, in the neighborhood of 3.5% annual for a very select sample for which we could do in just about any decade. 1976 was a Ford year. 1980 was a Carter year, with a recession -- if included 1977 to 1980, Carter's 4 year average would be around 8.6%/4 = 2.2%.

      Delete
    2. This is a bit of cherry picking isn't it? Even if you ignore the high inflation, those years look pretty bad if you look at the bigger picture. Real GDP growth (annualized) according to the BEA:

      1974 -0.5
      1975 -0.2
      1976 5.4
      1977 4.6
      1978 5.4
      1979 3.2
      1980 -0.2
      1981 2.6
      1982 -1.9

      Delete
    3. Real GDP factors out inflation. Your numbers show an average annual growth of just over 2%, which is better than the '00s under Bush/Obama which were under 1% and the 2010s under Obama which are just over 2%.

      You have a very selective period which includes three recessions including 1982 which was our worst post-WWII (after the '40s) recession until the Great Recession of 2008.

      The 1970s (1970 to 1979) had a Real GDP (2009 $s) growth of 36.94% for that 10 year period.

      Delete
    4. I might add that if you did not factor out inflation for 1970 to 1979 the annual growth would be well over 10% per year (I think something like 14%) -- the GDP more than doubled in this period.

      Delete
    5. Vic Volpe,
      Yes, I'm aware that real gdp is adjusted for inflation. Maybe I didn't word my post very well. My point was that the previous commenter is doing some major cherry picking when he keeps bringing up how much gdp grew by from '76-'79 as proof of the great success of inflation and money printing. By expanding the time frame a few years in either direction, those numbers start to look much worse. Still not as bad as recent years, but that's a different story.

      Delete
    6. This was almost entirely because of the entry of women and boomers into the labor force.

      Delete
    7. Like I said above, if it were women and boomers you would have seen more productive decades in the '80s and '90s -- these decades had more women and boomers (and Gen Xers) than the '60s. The numbers for the greater productivity of the '60s isn't even close -- on a GDP (real GDP, 2009 $s) per capita, the '60s are about 40% more productive than the '80s and about 70% more productive than the '90s.

      Delete
  16. dear prof Cochrane
    for your graphs, is the Y axis real gdp/time period, or real gdp capita-1/timeperiod ?
    cause if the population is growing rapidly, doesn't that often increase gdp ?
    and we are an aging (except for immigrant) population ?

    My understanding is that real gdp per capita has been growing at a truly astonishingly constant rate, about 2% year, for as far back as data goes, eg
    the url below is pretty typical (I think this graph, or a close cousin, was once on G Mankiw's blog)

    http://aneconomicsense.com/2012/07/20/the-shift-from-equitable-to-inequitable-growth-after-1980-helping-the-rich-has-not-helped-the-not-so-rich/

    ReplyDelete

Comments are welcome. Keep it short, polite, and on topic.

Thanks to a few abusers I am now moderating comments. I welcome thoughtful disagreement. I will block comments with insulting or abusive language. I'm also blocking totally inane comments. Try to make some sense. I am much more likely to allow critical comments if you have the honesty and courage to use your real name.