Why GDP lies about the true cost of a disaster

After a disaster, GDP can rise. Not just hold steady: rise. This paradox is no accounting curiosity. It is the sign that our main measuring instrument misses what matters most, starting with those who suffer the most.

Every year, disasters destroy some $327 billion of assets worldwide: homes, roads, factories reduced to nothing in a matter of hours. You would expect such devastation to show up in the headline economic indicators. And yet GDP, the figure everyone watches, does not fall in proportion. Worse: in the months that follow a disaster, it tends to rise. How can massive destruction coincide with growth that holds, or even accelerates?

This paradox has very concrete consequences. As long as we measure the cost of a disaster by GDP alone, we get both its scale and the policies meant to address it wrong, and we render invisible those who pay the heaviest price. Here is why.

The answer: GDP counts flows, not stocks

To understand it, recall what GDP measures: a flow of production over a period, not a stock of accumulated wealth. A disaster destroys, first of all, a stock: houses, schools, roads, equipment. That destruction is not subtracted from GDP, because a destroyed stock is not a flow. By contrast, everything spent to rebuild is counted as new activity. The result is that destruction goes unnoticed in the accounts while reconstruction is recorded as a plus.

This is not a theoretical detail. Economists have documented it for a long time: GDP tends to rise in the periods immediately following a disaster, precisely because the national accounts do not measure the stocks of capital destroyed, while they do measure the effort to replace them (Skidmore and Toya, 2002). An indicator that rises can therefore perfectly well conceal a real impoverishment.

What GDP doesn’t see

The first blind spot, then, is the destroyed capital itself. The wealth that vanishes with homes and infrastructure appears nowhere in a measure of flows.

The second is intangible losses: trauma, interrupted schooling, the erosion of social and community ties, environmental degradation. None of this fits into a box of the national accounts, and yet it is often what weighs most heavily on a life.

But the deepest blind spot lies elsewhere. GDP is an aggregate: it adds up. It therefore treats in the same way a dollar lost by a family on the edge of subsistence and a dollar lost by a wealthy household. Yet these two dollars are not worth the same. That is the whole point of shifting to a different measure.

Measuring what really matters: well-being

Economists have built a better lens. Rather than stopping at production, welfare economics asks the question of the real cost to the societies that bear the shock. Hallegatte and his coauthors have proposed its most complete application, tracing the full chain that links the physical hazard to the loss of well-being: exposure, asset destruction, income losses, consumption losses, and finally well-being losses. At each link, what matters is not only the average, but who is hit and with what capacity to absorb it.

The decisive idea fits in one sentence: a dollar of loss does not do the same harm depending on who bears it. For a household near the subsistence line, the same loss bites far harder, because its livelihood rests on fewer assets, its consumption is already stretched thin, it has no savings to cushion the shock, and it will take longer to rebuild. Economists call this the diminishing marginal utility of consumption, and it is precisely what aggregation into GDP erases.

From this follows a fairer measure. Hallegatte and his coauthors define socioeconomic resilience as the ratio of asset losses to well-being losses. When the two coincide, society spreads the burden of the shock perfectly; when that resilience falls, well-being losses exceed asset losses. Applied to 117 countries, the framework yields an average global resilience of 62 percent. In concrete terms, that means that for every dollar of assets destroyed, the loss of well-being amounts to about $1.60. But that average masks considerable gaps, from 25 percent in Guatemala to 81 percent in Denmark: for the same asset losses, one and the same disaster costs far more, in well-being, in the societies least equipped to absorb the shock. The $327 billion destroyed each year that I mentioned at the outset is therefore only the visible tip: the true cost, measured in well-being, reaches the equivalent of a $520 billion drop in annual consumption. That gap of nearly $200 billion is exactly what neither GDP nor damage accounting can see.

The real burden falls on the poorest

Why this gap? Because losses are not distributed at random. The same framework, applied globally, gives a result that should be far better known: the poorest 20 percent bear only 11 percent of asset losses, but 47 percent of well-being losses. In absolute terms, the poor lose less, since they own less. In well-being terms, that is, in terms of what the loss actually changes in their lives, they lose far more.

This imbalance comes from a triple disadvantage: poor households are more exposed, they lose proportionally more, and they receive less support to recover.

Prices add a further channel, especially sharp in an import-dependent island economy, as I discussed in the previous post. Alleyne and Lugay show that, in the Caribbean, the higher inflation climbs, the more relative prices scatter: some goods surge far faster than others. Here again, those who spend the largest share of their budget on food and energy take the hardest hit, and none of it shows up in a growth figure.

Together, these factors turn a disaster into a mechanism that deepens inequality. On a human scale, the order of magnitude is striking: disasters keep roughly 26 million people in extreme poverty at any given time. This is what a rising GDP can conceal: a society that has become both poorer and more unequal.

Change the instrument, change the policy

None of this is merely a quarrel about measurement. If we judge a disaster by GDP alone, we misjudge the cost, but also the cure. The welfare lens points to other levers: not only prevention ahead of the shock, but the capacity to absorb it afterwards, that is, social protection, financial inclusion, income diversification. These are policies that do not reduce the hazard, but that radically change what it costs the most vulnerable.

And that transforms a question we thought was simple: if the cost is so badly measured, what does “getting back to normal” really mean? That is precisely what I will examine in the next post, showing why the promise of a full recovery after a disaster is largely a myth.

References

Alleyne, D., & Lugay, B. (2011). Inflation and the Variability of Relative Prices in the Caribbean: Evidence from Panel Threshold Models. ECLAC, Project Documents, LC/W.444.

Hallegatte, S., Vogt-Schilb, A., Bangalore, M., & Rozenberg, J. (2017). Unbreakable: Building the Resilience of the Poor in the Face of Natural Disasters. Climate Change and Development. Washington, DC: World Bank.

Skidmore, M., & Toya, H. (2002). Do Natural Disasters Promote Long-Run Growth? Economic Inquiry, 40(4), 664-687.