26 July 2012

Comparing 2012 Drought Costs to 1980 and 1988

The enormous drought scorching the central USA will almost certainly cost at least $12 billion, making it the costliest since 1988, experts said Wednesday.

"There does seem to be near-unanimous agreement from industry experts that this year's drought losses will surpass the $12 billion recorded in 2011," says meteorologist Steve Bowen of Aon Benfield, a global reinsurance firm. . .

About 64% of the contiguous USA is in a drought, according to today's U.S. Drought Monitor, a federal website.

"Right now, it is difficult to say whether we end up reaching the loss levels of 1988 ($40 billion) and 1980 ($20 billion), given that it will be several months for agricultural industries to fully assess the total extent of their losses," Bowen says.
As readers here well know, comparing aggregate loss numbers over time requires understanding that context changes. It is not enough to simply adjust for inflation.

The graph at the top of this post shows the 1980, 1988 and 2012 (to date) estimates for drought damages, scaled to GDP (data from BEA), with 1980 scaled to 100. The graph shows that the 1988 drought was about 10% more costly than the 1980 drought, using the estimates reported by USA Today (which apparently derive from NCDC), even though it cost twice as much in current dollars.

At $12 billion the 2012 drought would be about 10% of the cost of the 1980 drought and less than that when compared to 1988. The costs of the 2012 drought are sure to rise in coming weeks and months, but they have a long way to go to exceed the standards set in 1980 and 1988.

So if you think 2012 is bad, you should know that it could be much, much worse.

UPDATE: I've been asked to show the actual values as a percentage of GDP. Here they are:

1980 = 0.72%
1988 = 0.78%
2012 = 0.08% (at $12 billion and counting)


  1. The initial assessment would suggest the effects are localized and manageable. Perhaps local resource availability is no longer the principal determinant of losses. Is this due to improved infrastructure, innovations (including technology), or distributed production?

    Well, as you have stated, this is a preliminary report. It would be interesting to learn what mitigating measures have been effective in preventing and overcoming losses caused by natural phenomenon.

  2. I am sorry, but this makes no sense whatsoever.

    If you want to compute welfare impacts, do a proper npv analysis.

    If you want to get a sense of how hard the agricultural sector is hit, look how much of gdp in each year came from ag, and then express the loss as percentage of that.

    The only thing you are picking up with your numbers is that other sectors of the economy expanded a great deal since the last droughts. While agricultural production grew since the 80s in absolute terms, its share in gdp is smaller because other sectors grew even faster.

  3. -2-dave

    Thanks for your comments.

    This post is not about npv of dollars, or losses within the ag sector or anything other than the size of estimated drought losses at 3 points in time as compared to contemporaneous GDP -- which is a perfectly valid metric, and is used all the time and in many contexts.


  4. Roger,
    Can it be that the $2 Billion is nowhere near the likely cost? If so, it dilutes your point, does it not?

    In what ways could the $2B number have been calculated short? Costs (losses) not yet realized, for example.

  5. RPJr- So Wikipedia says the agricultural sector in 2011 was 1.2% of GDP. Presumably it's shrinking over time, so soon it will be impossible to have a drought as bad (as a percentage of GDP) as 1980 or 1988.

    So what useful information does your 'valid metric' convey to us? That soon it will be impossible to have a drought as bad as 1988? I still don't get it.

  6. -4- n-g

    Thanks ... If I tell you that a loaf of bread cost $0.05 in 1925 and costs $3.00 today would you say that it is 60 times more expensive? Of course not.

    Also, drought impacts go well beyond just the ag sector. Thanks!

  7. i recall my earlier question - dyslexia again.

    loss as percentage of gdp makes sense to me. 12 billion still seems low as an absolute cost in comparison to 1988 though. are we sure all the costs are realized?

  8. -7-j ferguson

    The costs will go higher, surely, but how high? About $120 billion rivals 1988...


  9. The picture doesn't look much different in you scale in terms of the CPI instead of GDP: