Thursday, November 1, 2018

Roger Pielke Jr's weather disaster essay is too simplistic, and befuddles deniers at WUWT

Sou | 4:28 AM Go to the first of 6 comments. Add a comment
It's a short article. Short in length and short on substance. I'm referring to a paper written by Roger Pielke Jr. where he attempts to report on whether and how much progress there has been in a small part of one of the seventeen United Nations' Sustainable Development Goals (SDG).

The paper attracted the attention of Anthony Watts, a science denier who runs a blog known as WUWT. Anthony, not being the brightest spark, not even in the dark deniosphere where the bar for brightness is low, got the paper upside down and inside out. More on that later.

Sustainable Development Goals

The UN's SDG has 17 goals aimed at improving societies, the well-being of people, and the sustainability of the planet. Each goal has several parts and, at present, 232 unique indicators. The indicators are for measuring progress toward achieving the goals.
Roger Pielke decided to address just one of the 232 unique indicators. He worked out direct economic losses worldwide as a proportion of global gross domestic product (GDP). The period he reported on was from 1990 to 2017. His published article was short and, if truth be told, it was little more than a blog post.

The SDG goals are mainly about improving the well-being of people, particularly those less well off at present. The report containing indicators led with the reaffirmation of the pledge in a typical mouthful of a sentence:
that no one will be left behind in implementing the  2030 Agenda for Sustainable Development, that the 2030 Agenda is people-centred,  universal and transformative, that the Sustainable Development Goals and targets  are integrated and indivisible and balance the three dimensions of sustainable  development – economic, social and environmental – and that it is a plan of action for people, planet and prosperity that also seeks to strengthen universal peace in  larger freedom, to be implemented by all countries and stakeholders, acting in collaborative partnership, and reaffirming further all the principles recognized in the Agenda and that eradicating poverty in all its forms and dimensions, including extreme poverty, is the greatest global challenge and an indispensable requirement for sustainable development

In his paper Roger referred to two goals:
  • Goal 1 is to end poverty in all its forms everywhere. Part 1.5 of Goal 1 is:
By 2030, build the resilience of the poor and those in vulnerable situations and reduce their exposure and vulnerability to climate-related extreme events and other economic, social and environmental shocks and disasters
  • Goal 11 is to make cities and human settlements inclusive, safe, resilient and sustainable. Part 11.5 is to:
By 2030, significantly reduce the number of deaths and the number of people affected and substantially decrease the direct economic losses relative to global gross domestic product caused by disasters, including water-related disasters, with a focus on protecting the poor and people in vulnerable situations
Notice that in both goals the emphasis is on people. There is a lot of focus on improving the lives of the poor and people in vulnerable situations.

The indicator 1.5.2 is "Direct economic loss attributed to disasters in relation to global gross domestic product" and this is all that Roger looked at. He also mentioned the indicator 11.5.2, however, that one is about more than global GDP, it includes:" Direct economic loss in relation to global GDP, damage to critical infrastructure and number of disruptions to basic services, attributed to disasters". Roger didn't address the issues of infrastructure damage and service disruption.

In short, Roger took a very small sliver of a sliver, or one of the 232 unique indicators, which is fine in itself but not all that useful in the context of the SDG goals in my view. At least, not without a broader contextual analysis and discussion of impacts on the poor and vulnerable.

Indicators are not intended to be stand-alone. They are generally assessed in conjunction with each other. Drawing conclusions about the success or otherwise of a goal based on just one indicator of a sub-goal among 17 goals and 232 indicators is not wise or terribly useful. Even with that one indicator, it would perhaps be more useful to look at economic losses in different regions.

The results - direct economic losses from weather disasters have increased by 74%

Let's leave aside the lack of depth and context for the moment and consider what Roger found. He discovered that weather related losses have increased by 74% since 1990. That number is based on a linear trend of reported direct economic losses normalised to 2017 US dollars.

Roger attributed the 74% increase in direct economic loss to increased population. He wrote:
The primary factor driving the overall increase in disaster losses is societal, mainly growth in populations and settlements at risk to the consequences of extreme events (IPCC, 2012). While some weather and climate extremes are expected to increase in frequency and/or intensity in the future, to date there is not strong evidence of such increases in tropical cyclones, floods, drought or tornadoes on climate time scales (IPCC, 2018; Pielke, 2018). Of course, any calculation of trends in catastrophe losses is sensitive to choice of start and end date, so caution is urged in their interpretation. 
Population expansion would certainly be a factor. Oddly, he made no mention of any impact of stronger building codes or the increased wealth suggested by the increase in GDP he observed. I expect there'd be climate researchers who would dispute the implication about evidence of weather extremes and disasters, too.

An oddly simplistic linear analysis

In terms of disasters related to weather and climate as a proportion of GDP, Roger said he found "a declining trend during the period 1990 to 2017, from more than 0.25% of global GDP to below 0.20%". He based that on the trend of a red line :)

Below is a chart from Roger's paper. I've added text from his paper to help explain it.

Figure 1 | Weather disaster losses as a percentage of global GDP. Figure 3 in Pielke18

Another simplistic perspective

Well, I suppose that's one way of looking at it. How about another. The chart below is not in Roger's paper. I prepared it myself using data from Table 1 in Roger's paper. I've just included the Munich Re data with global GDP. No percentages or proportions, just plain numbers. The data is grouped in 5 year periods. The blue columns are GDP (left axis) and the dark brown marks are weather disaster losses in 2017 $US (right axis).

Figure 2 | Global GDP and direct economic losses from weather-related disasters as recorded by Munich Re. Data source: Table 1 in Pielke18
What do you think? I'd say the spurt in GDP growth from 2010 to 2014 is largely China. It may be leveling off - too early to tell. The period from 1995 to 1999 was a bad one for weather events in terms of economic loss, but only in relation to the five years before and after. Then there was a jump up. The losses for the current five year period are already piling up with a vengeance - they are the biggest of all so far. (It will get worse.)

There is a wealth of information yet to explore

Munich Re gives a much more detailed view of weather disasters, reporting them by income level by country, which has more meaning in terms of the impact on societies. It also has a report on how they work out economic losses from disasters. A lot of thought seems to have gone into this, as you'd expect. For example, they talk about allowing for differences in impact between regions:
Loss data normalisation makes historical loss values comparable over time. However, loss data normalisation does not take into account that a loss of, for example, US$ 100m is of quite different significance in Haiti or Bangladesh than in Germany or the USA. To be able to compare the impact of loss events on different countries, we need a severity or impact classification scheme that takes these differences into account.
No issues such as that were discussed in Roger's paper. Perhaps he'll do another deeper analysis - or maybe he'll leave that to an economist or social scientist.

Aggregate GDP is not very useful as a comparison with weather-related losses

Although Roger mentioned the increase in direct economic losses from weather disasters, he didn't mention the size of the overall increase in GDP. The chart below is from the source Roger had in his paper, the Federal Reserve Bank of St Louis:

Figure 3 | World GDP 1990 to 2017. Federal Reserve of St Louis.

World GDP rose from $US22,573,777,725,713.20 in 1990 to  $US80,683,787,437,857.80 in January 2017. GDP in 2017 is 3.6 times that in 1990. Is it any wonder that despite the rise in disasters, they weren't associated with a drop in proportion to GDP? All that can be said is that, over that period, GDP rose more quickly than the direct economic cost of weather disasters worldwide.

I've always disliked using GDP as a yardstick by which to measure the impact of weather disasters. I'm not alone. For one thing, recovery and reconstruction after disasters can drive an increase in GDP, even though they cause a loss of income, assets, wealth and welfare to individuals and localities. Also, the impact on a region will vary, depending on the capacity to recover (compare Houston to Puerto Rico), and how often disaster strikes.

There's more. In Chapter 2 of S. Hallegatte, Natural Disasters and Climate Change, there was discussion of mechanisms that determine the economic consequences of disasters and welfare impacts.  The chapter highlighted some counter-intuitive findings, such as:
...economies in recession are more resilient to the effects of natural disasters. This result appears consistent with empirical evidence. For instance, the 1999 earthquake in Turkey caused direct destructions amounting to 1.5–3 % of Turkey’s GDP, but consequences on growth remained limited, probably because the economy had signi´Čücant unused resources at that time (the Turkish GDP contracted by 7 % in the year preceding the earthquake). In this case, therefore, the earthquake may have acted as a stimulus, and have increased economic activities in spite of its human consequences. In 1992, when hurricane Andrew made landfall on south Florida, the economy was depressed and only 50 % of the construction workers were employed (West and Lenze 1994). The reconstruction needs had a stimulus effects on the construction sectors, which would have been impossible in a better economic situation.

The article also addressed one of the main things I found lacking in Roger's paper, how the different impact in different places can be hidden by aggregating numbers. This is what his numbers don't show.
GDP and aggregated consumption can be misleading instruments to measure actual welfare losses, especially on the poorest. This is particularly important for disasters, as the poorest are usually more exposed and less able to protect themselves and recover from shocks. As already mentioned, Rodriguez-Oreggia et al. (2009) show that municipalities affected by disasters in Mexico see an increase in poverty by 1.5–3.6 % point. Baez and Mason (2008) present evidence for El Salvador, where earthquakes in 2001 reduced income and consumption per capita of the most affected households by one third. 

Poor old Anthony Watts doesn't understand plain Rogerian English

I'll finish with the ridiculous, which came not from Roger Pielke Jr, but from Anthony Watts at WUWT. You'll recall how Roger reported an increase in direct economic loss of 74% (albeit by very simplistic analysis). Well, look at what happened at WUWT.

Anthony Watts' headline was the opposite to what Roger reported: New paper by Pielke Jr. – weather related disasters are decreasing in costs

No, Anthony. Roger showed they are increasing in cost, not decreasing. There's no excuse. Anthony even copied and pasted the abstract, which included this sentence (my emphasis):
In constant 2017 US dollars, both weather-related and non-weather related catastrophe losses have increased, with a 74% increase in the former and 182% increase in the latter since 1990. 
Just another example of Anthony Watts not reading what he posts at WUWT - or worse, reading it and not understanding it - or still worse, deliberately lying.

References and further reading

Roger Pielke (2018) Tracking progress on the economic costs of disasters under the indicators of the sustainable development goals, Environmental Hazards, DOI: 10.1080/17477891.2018.1540343

World Bank GDP data - with interactive charts

Not if, but how - NatCatSERVICE Natural catastrophe know-how for risk management and research - interactive charts and tables on natural catastrophes, provided by Munich Re

Not If but How - NatCatSERVICE Methodology, Munich Re, March 2018

What Is a Disaster? An Economic Point of View. Chapter 2 in S. Hallegatte, Natural Disasters and Climate Change, DOI 10.1007/978-3-319-08933-1__2 (pdf here)

The Sustainable Development Goals Report 2018, United Nations New York

Resolution adopted by the General Assembly on 6 July 2017 71/313. Work of the Statistical Commission pertaining to the 2030 Agenda for Sustainable Development - the paper including all the SDG goals and indicators

From the HotWhopper archives


  1. Welcome back.
    Roger Pielke Jr's spin is easily debunked by reference to Munich re's graph of natural disasters https://www.iii.org/sites/default/files/graphs/num_world_nat_cats_80-17.gif
    If weather costs were truly not being effected by climate change the cost of geophysical disasters would rise in tandem with weather disasters.
    They dont geophysical disasters are near static while weathers costs have increased considerably strongly suggesting that building codes and rising wealth counteract the increasing value of assets in harms way.

    1. Absolutely, Kiwi Griff. I'll point out in his paper he said he expects weather disasters to increase with global warming. On the other hand, he also tries to downplay the impact of climate change, as seems to have been the reason for the slant he took in this paper.

      I was going to raise the question of where Roger put hydrological events - were they included with "weather and climate" disasters or lumped in with geological disasters under "overall disasters".

      I expect each hydrological disaster would need to be analysed to determine the root cause. E.g. was a disaster caused by faulty dam construction or was it caused by extreme monsoonal rains (or both or neither)?

  2. yah! Sou returns. Time to dust off the popcorn maker.

  3. Population growth is a key factor that you seem to dismiss.
    Meanwhile Goal 1
    In 2000 the United Nations set a goal of eradicating poverty by 2030. There is general agreement that the proportion of destitute people in world has been reduced by much more than 50 percent already.
    And there is little doubt Global average life expectancy increased by 5.5 years between 2000 and 2016, the fastest increase since the 1960s.

    1. You must have missed this, Jeff:

      "Population expansion would certainly be a factor."

      It's under Roger's quote about the factor he attributes to the increase in direct economic costs.

  4. I do love the idea that, here in the UK, increasing damage due to flooding in towns across the country is caused by increasing profits in the City of London.


Instead of commenting as "Anonymous", please comment using "Name/URL" and your name, initials or pseudonym or whatever. You can leave the "URL" box blank. This isn't mandatory. You can also sign in using your Google ID, Wordpress ID etc as indicated. NOTE: Some Wordpress users are having trouble signing in. If that's you, try signing in using Name/URL. Details here.

Click here to read the HotWhopper comment policy.