The Disaster Profiteers: How Natural Disasters Make the Rich Richer and the Poor Even Poorer

by John C. Mutter

  Extreme and growing social inequality is the source of countless social ills and of financial disaster. It is one of the great challenges of our time. Removing the opportunities for profiteering from disasters will not only right an injustice but also help to bring us closer together.

  Technical Appendix I

  Simplified Socioeconomics of Natural Disaster Shocks and Their Consequences

  This appendix amplifies some of the issues raised in chapter 1. Consider Figure 1.1

  Figure 1. The transformation of a physical shock into a social shock.

  On the left is a generic natural shock represented in diagram form. It could be an earthquake because it looks a little like the swing of a seismograph needle, but it doesn’t have to be. It is characterized by the time of onset of the event, Te (when the event starts), and its magnitude, A.

  On the right is a diagram of the social consequence of the shock. The horizontal axis is time, and the vertical is some measure of welfare. It could be gross domestic product (GDP), but it could be any one of many others discussed in chapter 1, such as the Human Development Index (HDI) or Gross National Income (GNI).

  The disaster event occurs at the time designated Te, the same as on the left. The heavy line is the growth in welfare that we assume would have occurred had the disaster not happened. It is constructed by taking the growth before the disaster and simply extrapolating on past Te. We can think of it as the growth rate unperturbed by disasters.

  At Te there is an immediate loss of something by the amount L. What it usually means is the replacement cost of losses of capital stocks such as homes and businesses, roads and airports. It may be people, but what it usually refers to is some sort of loss of a manufactured capital. Remember that this sort of built capital isn’t directly included in the GDP formula in chapter 1, so GDP might not show an immediate drop.

  What follows is the recovery, designated R. It is shown in the diagram as if the social welfare resumes growing immediately after the disaster at the same rate as it did before, but starting from the lower level (the level it was at just before the disaster, minus L). Although that is quite unrealistic because you would reasonably expect the economy to do less well after being hit by a disaster, it sets the stage for our discussion.

  Two things about the expected rate of growth in a disaster-impacted society are important and are illustrated in Figure 2. The horizontal and vertical axes are as before.

  Figure 2. The effect of the same loss on economies at different levels of development and growth rates.

  One important factor is the slope of the line that describes the rate at which welfare is improving. Some economies are growing slowly, some quickly. The lower line has a much steeper upward slope than the upper line, so it describes a society that is improving in welfare faster than the upper line—over the same period of time, welfare increases by a greater amount for the lower line.

  The other thing to note is that the absolute level of welfare is different in the two cases as well. The upper line lies in a higher part of the graph, so the absolute level of welfare at all places along that line is greater than for the lower line. So the lower line describes a fast-growing economy but one that has yet to achieve a high absolute level of welfare. The upper line might represent a country more developed than the lower one. These differences are consistent with empirical evidence that developed economies tend to grow more slowly than economies of developing nations, such as China and India, which have seen double-digit growth.

  The amount of the disaster loss (the downward arrow) is shown as the same absolute amount in both cases, and it is fairly obvious that the same loss in these two economies might have different effects. For the upper line, the loss is a smaller percentage of the total welfare than for the lower case.

  What constitutes recovery? Figure 3 shows two cases; the one to the left is for a relatively fast-growing economy and the one to the right is for a relatively slow-growing economy. Each experienced the same absolute loss. The times denoted T1 and T2 are the times at which welfare has returned to the level it was at before the disaster.

  T2 is clearly much greater than T1, so the expectation is that the faster-growing country will recover to its predisaster level more quickly than the slow-growing country. The absolute level from which the disaster fell is not important here. This slow recovery time is something that greatly concerns the World Bank when it considers what sort of actions to take after a major disaster in poor countries that are not growing very rapidly. What is also clear from this simple diagram is that the larger the drop, the longer the recovery time will be and that even a small drop in a slow-growing country can require a very long time for recovery, just because the natural growth rate is slow.

  Figure 3. The effect on recovery time due to growth rate.

  As measured in GDP, the US economy has returned to where it was before the recession. If the vertical axis was employment, things would not look different.

  It is very reasonable to suppose that there would be a period of slow or zero growth after the disaster, and perhaps even a period of negative growth, meaning continued losses and retraction of the economy. That is illustrated in Figure 4.

  Figure 4. The effect of a period of post-disaster stagnation on recovery time.

  The period of stagnation is designated S. The natural growth rate and the size of the loss, L, are the same in both diagrams. What is intuitively clear is that any period of stagnation lengthens the time to recovery by the length of the stagnation period—in other words, T1 is extended by S1, and T2 is extended by S2. What is perhaps less obvious is that stagnation periods, in effect, increase losses—L2 and L3 are larger than the original loss, L. Not only capital is destroyed; also destroyed is the welfare that would have accrued during that period had the disaster not happened.

  You can think about these graphs describing disaster losses and recovery in much the same way as you think about the 2008 financial crisis and recession and the recovery after it. In June 2014, the New York Times published a figure very similar to Figure 4, soon after the US government declared that the economy had finally reached its prerecession level.2 The drop in GDP wasn’t as precipitous as shown in the disaster graphs because the recession, although evolving alarmingly quickly, took time to fully develop.

  Another way to extend the recovery time, though it may seem counterintuitive, is to make use of national reserves, as shown in Figure 5. These reserves are denoted R and r in the diagrams, where R is larger than r.

  Figure 5. The effect of the use of capital reserves.

  The use of national reserves can have the effect of increasing the loss. In doing so, it increases the recovery time. That is because the initial loss, L, as well as the amount of reserve that has been drawn on to help effect recovery also have to be regained. The capital that is represented by R or r is financial capital. It is not the same as the capital lost in L, which is built or manufactured capital, so it’s really not correct to just add it on to L as I have done, but it illustrates the point. Financial capital will always be needed for recovery, but it is much better for many countries, especially poorer ones, if that capital comes from outside in the form of direct aid or a grant than if it has to come from the countries’ reserve funds or loans because using reserves or loans also adds to a country’s debt and hence requires repayment.

  Drawing on reserves might be thought of as reducing losses, something that would make L smaller by shaving some losses off the bottom. But what is lost is lost. Expenditures to counter those losses are also lost in the sense that they are no longer available for other, more productive uses, and they also have to be replenished.

  Whether the use of capital reserves has a positive or negative effect very much depends on what the reserve capital is spent on. If it is spent on temporary housing, cleaning up debris, or emergency health care, it is really lost. If it can be used
to restore infrastructure, it could have a positive effect on economic recovery. One of the best uses of reserves is to reduce the stagnation period by getting things moving again quickly after a disaster.

  Returning to an earlier question: What do we mean by a disaster loss? Figure 6 repeats one that we used earlier. The only difference is that we have filled in a triangular area.

  The shaded area is the time integration of the losses accumulated from the disaster instant until the point at which recovery to the predisaster level of welfare has been achieved—T1 in the faster-growing case, or T2 in the slower setting. This integration of losses is much closer to expressing the true economic loss than the initial loss, L, described earlier.

  The area of the lower triangle is larger than the area in the upper one even though the initial losses are the same.3 That means that a slow-growing country could lose as much as a fast-growing country, even if it suffered smaller initial losses. So the slow-growing country could take longer to recover and have greater overall losses.

  But the points on the recovery line are still below the counterfactual line—the line where the economy would have been had the disaster not happened. What recovery should mean is a return to the initial growth line, not the predisaster level of welfare. Figure 6 shows this.

  Figure 6. The shaded area may be a better way to express disaster loss.

  The economy was at point A before the disaster. At point A', the economy has returned to the initial level A. But, had the disaster not happened, the economy, ceteris paribus, would have grown to be at B.

  Thinking back to the US economy and its recovery from recession, in June 2014, the US Labor Department announced that the economy was finally employing as many people as it had been before the recession in 2007. So in this diagram, you can think of the vertical axis as jobs. In the US economy, we are at A'. What the Labor Department announced was, in essence, that there has actually been no growth in employment in six years; catching up the amount lost cannot be thought of as growth.

  The economy is growing again and has more or less reached its 2007 level, but it is nowhere near a projection of where it would have been had pre-2007 growth continued.

  On the day of the announcement, the New York Times online edition produced an analysis of the recovery in 255 interactive graphs.4 The purpose of the graphs was to show that the recovery has been very heterogeneous as measured in employment. The oil and gas sector has recovered to where it would have been had the recession not occurred—back to point B on our stylized plot in Figure 7. This may have been helped by the shale gas revolution.

  Figure 7. Illustration of recovery to predisaster level and predisaster level plus expected growth.

  Furniture stores, however, were not doing particularly well before the recession. They were hit very hard in the crisis and have not recovered their losses at all. The business, environment, and other consulting area has recovered to its prerecession level but not to where it would have been without the recession. So the businesses growth trajectory has recovered but not the absolute level of jobs.

  The graphs in the New York Times show clearly that different sectors of the economy have experienced job recovery at different rates, including what might be called a negative rate. Natural disasters will affect different businesses and different people differently. In a very broad analogy, perhaps we could think of Haiti after the 2010 earthquake as the furniture business. Perhaps the consulting business is like New Orleans after Hurricane Katrina. It’s harder to find an example to match the oil and gas industry.

  So, in a stylized way, what we really want to achieve is shown in Figure 8.

  Figure 8. Full recovery occurs when the line representing recovery meets the line on which the economy would have grown had the disaster not happened.

  Now the postdisaster growth line has met the expected growth trajectory and a different, larger triangular area is involved. What you notice right away about this is diagram that in order to meet the predisaster growth trajectory, the recovery growth line has to slope upward more steeply than the natural growth trajectory. That means that if an economy is ever to return to where it would have been without the disaster, it has to grow faster in the postdisaster recovery period than it did before.

  In Figure 9, the two counterfactual growth lines are the same.

  Figure 9. Full recovery at different recovery rates.

  What is different is that the initial loss in the upper example is greater than in the lower example, but the rate of recovery is faster in the upper one than in the lower one. The area in the upper example is clearly smaller than in the lower example. The upper case can be thought of as an economy capable of rapid response even in the face of large absolute losses. It more likely represents a well-developed country; the lower line represents a poorer situation.

  Could there be a silver lining? Imagine that postdisaster growth was somehow made faster than in the predisaster period. A massive infusion of capital from outside the economy with no debt obligation and spent exceedingly well might achieve that. Then why should that slope of high growth stop when the natural growth line has been reached? Figure 10 illustrates the idea.5

  What you have to imagine is the following scenario. In almost any urban setting, there will be old and new homes, new and old buildings, new and old bridges and roads and ports and other critical facilities. Images of cities in developing countries often show slums on the periphery, luxury high-rises in the center, and perhaps a beautiful beach behind.

  In almost any type of natural disaster in a developing country, practically everything is damaged to some extent, but some structures remain fairly intact because, by intent or happenstance, they are stronger. It is reasonable to expect that the older, weaker capital stocks, including critical infrastructure, will be damaged far more severely than the newer capital stock. The old capital would normally be replaced with new capital that is better and more efficient in every way, if only because it’s newer. Old roads and bridges are replaced with new wider, stronger roads and bridges. New schools and hospitals appear. The economy with newly replaced infrastructure critical to commerce should run faster than before. A forced technology upgrade led to greater growth.

  Figure 10. Recovery overshoot.

  Figure 11 illustrates how a two-component economy might be affected by disaster.

  Figure 11. Full recovery in two sectors of society and the growth of inequality.

  Everything is as before except that there are two lines of predisaster growth, the upper one representing wealth and strong growth, the lower one representing the lower overall assets and slower growth. The growth rate on the lower line is smaller and the absolute level is lower. Just before the disaster, the inequality between the two is d and is widening. A disaster strikes as before. Now we see that the drop from the upper curve is much greater than that from the lower curve, but recovery is much faster. The lower sector loses less, but the recovery is much slower or perhaps not at all.

  Now it is easy to see that after a short time, the gap between the two new growth curves has increased—D is now greater than d and is greater than what the gap would have widened to without the disaster. Inequality was sure to have increased anyway, but now it has increased all the more. The disaster has exacerbated inequality. But it’s not the immediate effect of the disaster that matters, it’s the effect of differing recoveries: A rich, capital-owning group appears to have lost a lot more than the poorer group, but in the end it gains at the expense of the poorer group.

  TECHICAL APPENDIX II

  Disasters in Neoclassical Growth Theory

  Neoclassical growth theory is attributed to Robert Solow, and the ideas are often referred to as the Solow model, or the Solow-Swan model.1 (Swan refers to T. S. Swan, who came up with almost the same model independently of Solow, both in 1956.) The ideas are contained in Figure 1.

  The
vertical axis is the output per person, usually designated q, but other letters also are used (y is common). Output per person is created by a so-called production function,2 here written as Af(k), where A is the total factor productivity and k is the capital-labor ratio. So the horizontal axis is capital, and the curve shows how output depends on capital. That means that output is driven by capital (strictly, capital in relation to available labor). The upper curve that plots that function sometimes is called an L-curve for its shape. Near the origin (lower left corner), the curve is steep, meaning that for small additions of capital, large increases in production occur. It is almost flat in the upper right so that additions of capital achieve less; there is a diminishing return on marginal capital investment.

  The lower curve, which is very similar, is the same function with s as the factor modifying the first curve, sAf(k), where s denotes the savings rate. Because no one ever saves everything, s is less than 1, and the second curve will always sit below the first. Then there is a straight line (n + d)k. Here k is the same as before, n is population growth rate, and d is the rate of depreciation of capital.

  Figure 1. A standard depiction of the Solow-Swan exogenous growth model. All symbols are explained in the text. Reference 1 gives citations for the model.

  The term (n + d)k is called capital widening; it is the amount of savings necessary to keep the capital-labor ratio the same despite depreciation and population growth. Those latter two elements can be thought of as having similar effects on growth—if the population increases greatly but the capital stock stays the same, then production per capita will drop because the model regards capital as the source of growth. In the opposite case, where the capital depreciates and the population is the same, productivity decreases as well. There is a point where this straight line meets the savings curve, sAf(k); that is the point when the economy is in equilibrium. At that point, per person capital is designated kE. We could associate kE with an equilibrium output qE. In the standard theory, when the capital-widening line is below the savings curve (perhaps a more realistic way to say it is that the savings curve exceeds capital widening), the economy grows. That is, output grows.