The analysis in this study uses current production (or population) for country weights in obtaining global estimates for the impact of climate change on agriculture. This approach tends to understate the future losses, because it is likely that by late in this century the business as usual baseline for agricultural output would have shown a much larger share for developing and low latitude countries than is the case at present. So from this standpoint the losses may be understated.

In contrast, it might be argued that dynamic considerations will typically shrink the relative importance of losses from climate change, because technological change can be expected to raise yields by far more than global warming reduces them. Comfort from the prospect of rising yields from technological change may nonetheless fail to take into account the fact that rising demand for agricultural products may run a close race with technological change, so that yield losses to climate change could still do damage that more than exceeds any excess supply trends in the baseline. Moreover, an important additional factor must be incorporated into the dynamic analysis: the likely diversion of agricultural land to production of biomass for ethanol.

This chapter seeks to arrive at some ballpark estimates of the net effect of these divergent influences. Define the ratio of demand or supply in 2085 to the level in 2005 as X for each of several dimensions. Consider first population. The United Nations (2006) projects population in 2050 at 9.08 billion in its medium case, compared with 6.46 billion in 2004. Population grows at a pace of 0.42 percent per year in the decade 2040-2050 in this case. In its high case, the United Nations projects world population in 2050 at 10.65 billion, with annual growth at 0.93 percent in the decade 2040-2050. If these two respective levels and rates are used for projection to 2085, the resulting global population in 2085 is 10.52 billion in the

Figure 6.1 Income elasticity of demand for food, tobacco, and beverages and purchasing power parity (PPP) income per capita

PPP income per capita (dollars) 45,00040,00035,00030,00025,00020,00015,00010,0005,000

income elasticity of demand medium case and 14.72 billion in the high case. The expansion factor from current levels to 2085 for population is thus XN = 1.63 in the medium case and 2.28 in the high case. In broad terms, global agricultural output will need to double, approximately, to keep up with population growth over this period.

Also, demand will increase from rising per capita incomes. Figure 6.1 shows the relationship of the income elasticity of demand for food, beverages, and tobacco (as calculated by ERS 2006b) to purchasing power parity (PPP) GDP per capita (World Bank 2006) for 64 countries. There is a clear inverse relationship between per capita income and income elasticity of food, which amounts to a strong form of Engel's law (which states that food expenditure rises less than proportionately with rising income).1 If the regression equation relating the two is applied to global average PPP GDP per capita for 2004 ($6,329), the resulting global income elasticity for food at present is 0.655.2 Even if per capita income grew at 1 percent per year through 2045 (the midpoint of the period considered), the global income elasticity would still be relatively high at 0.612.

Assuming that per capita income grows at 1 percent annually over the next 80 years, and that the average income elasticity is 0.612, then rising

1. A weak form would be any income elasticity less than unity.

2. A simple regression yields the following results: 6 = 0.744 (100.1) - 1.4 X 10 -5 (-34.7) y*, adjusted R2 = 0.95 (f-statistics in parentheses), where 6 is income elasticity of food demand and y* is PPP per capita GDP in dollars.

0 0

Post a comment