Climate Change 2001:
Working Group II: Impacts, Adaptation and Vulnerability
Other reports in this collection

8.3.2. Risk Sharing between the Private and Public Sectors

The private insurance industry is part of a larger community that bears the costs of weather-related events (Ryland, 2000). The nature and cost of weather-related losses vary considerably around the globe, as does the portion of the loss that is privately insured. Private insurance pays a higher proportion of benefits for storm-related losses than for any other weather-related event, although flood insurance has a particularly low rate of coverage (Figure 8-6).


Figure 8-4:Trends in U.S. natural disasters, insurance industry profitability, and solvency, 1969-1999: (a) Sensitivity of property/casualty insurance sector net financial results to investment income and underwriting gain/loss. Upper bars indicate investment income; lower bars indicate net result of core business (premium revenues vs. claims paid). Curve is the net result. (b) Annual number of insolvencies and natural disaster losses (Mills et al., 2001).

Insurers bear only 20% of the total economic costs of weather-related events globally. The ratio is far lower in developing countries (e.g., 7% in Africa and 4% in Asia for the year 1998) (Munich Re, 1999b). Even in countries where insurance penetration is high, insurance can account for less than half of the weather-related payouts—for example, 27% in Europe, 30% in the United States, 34% in Australia (Munich Re, 1999b), and 20% in Canada (EPC, 2000). In a review of four major wildfire and flood catastrophes in Australia, Leigh et al. (1998a,b) found that private-sector insurers bore 9-39% of the total economic losses; a comparable amount was provided by local and federal governments. Other entities assuming such costs include federal disaster relief providers, local governments, and uninsured property owners (Pielke and Landsea, 1998)—as in the case of Hurricane Andrew, in which only half of the losses were insured (Pielke, 1997).

One important risk-assuming group, the corporate self-insurance market, is growing rapidly. In the United States, such premiums are approaching the level of the traditional commercial insurance market (roughly US$134 billion) (Best's Review, 1998; Bowers, 1999).

Where insurers will not or are directly or indirectly regulated not to accept specific catastrophe risks, governments in many countries—including Belgium, France, Japan, The Netherlands, New Zealand, Norway, Spain, and the United States—may adopt the role of insurer or reinsurer or of regulator in establishing risk-pooling mechanisms (III, 2000b). Programs in France, Japan, and New Zealand explicitly define the governments' role as paying for "uninsurable damages" (CCR, 1999; Gastel, 1999). In some countries (e.g., Canada, Finland, France, Norway, the United States) this is the case for drought or other agricultural risks, and in others (e.g., Japan) this is limited to earthquake risks. Such schemes can grow rapidly, as illustrated by the jump in the numbers of policies under the Florida Windstorm Underwriting Association from 62,000 to 417,000 between 1992 and 1997 (Anderson, 2000).

Government's role in providing resources for disaster preparedness and recovery and insurance products related to natural disasters also is a key moderating factor in insurers' involvement in such risks. It can be a two-edged sword: It provides a platform for private industry to participate, but it also can drive consumers away from commercial market solutions (Klein, 1997; Pullen, 1999a). The absolute value of government payments for natural disasters is poorly documented, and the statistical record is fragmented. The United States made disaster-related payments of US$119 billion (1993 US$) over the 1977-1993 period, equivalent to an average of US$7 billion yr-1 (Anderson, 2000). The Japanese government has devoted 5-9% of its national budget to disaster preparedness and recovery in recent decades (Sudo et al., 2000).

Flood insurance merits special mention, given the magnitude of risks and losses, the difficulty of establishing fair and actuarially based rates, and the connection between flood and climate change (see Chapter 4; Aldred, 2000). Recent analyses in the United States found that 25% of homes and other structures within 150 m of the coastline will fall victim to the effects of erosion within 60 years (Heinz Center, 2000). Sea-level rise will impact flood insurance through inundation and erosion resulting from storm surge (see Chapter 6). Countries differ widely with regard to their approach to defining and financing flood risks via private-sector (re)insurance versus public mechanisms (Van Schoubroeck, 1997; Gaschen et al., 1998; Hausmann, 1998). Hybrid public-private systems and government-only systems also can be found (e.g., in the United States), as can systems with no formal flood insurance whatsoever.

Table 8-3: Billion-dollar and larger insurance losses, 1970-1999, as of December 2000 (Munich Re, 2000). Figures are adjusted for inflation (1999 values).
Year
Event
Area
Insured Losses (US$M)
Economic losses (US$M)
Ratio of Insure d /
Economic Losses

1992

Hurricane Andrew USA
20,800
36,600
0.57
1994 Northridge earthquake USA
17,600
50,600
0.35
1991 Typhoon Mireille Japan
6,900
12,700
0.54
1990 Winterstorm Daria Europe
6,800
9,100
0.75
1989 Hurricane Hugo Caribbean, USA
6,300
12,700
0.50
1999 Winterstorm Lothar Europe
5,900
11,100
0.53
1987 Winterstorm Western Europe
4,700
5,600
0,84
1998 Hurricane Georges Caribbean, USA
3,500
10,300
0.34
1995 Earthquake Japan
3,400
112,100
0.03
1999 Typhoon Bart Japan
3,400
5,000
0.60
1990 Winterstorm Vivian Europe
2,800
4,400
0.64
1999 Winterstorm Martin Europe
2,500
4,100
0.61
1995 Hurricane Opal USA
2,400
3,400
0.71
1999 Hurricane Floyd USA
2,200
4,500
0.49
1983 Hurricane Alicia USA
2,200
3,500
0.63
1991 Oakland fire USA
2,200
2,600
0.85
1993 Blizzard USA
2,000
5,800
0.34
1992 Hurricane Iniki Hawaii
2,000
3,700
0.54
1999 Winterstorm Anatol Europe
2,000
2,300
0.87
1996 Hurricane Fran USA
1,800
5,700
0.32
1990 Winterstorm Wiebke Europe
1,800
3,000
0.60
1990 1990 WinterstormHerta Europe
1,800
2,600
0.69
1995 Hurricane Luis Caribbean
1,700
2,800
0.61
1999 Tornadoes USA
1,485
2,000
0.74
1998 Hailstorm, tempest USA
1,400
1,900
0.74
1995 Hailstorm USA
1,300
2,300
0.57
1993 Floods USA
1,200
18,600
0.06
1998 Ice storm Canada, USA
1,200
2,600
0.46
1999 Hailstorm Australia
1,100
1,500
0.67
1998 Floods China
1,050
30,900
0.03

A central question is whether changes in natural disaster-related losses will generate increased reliance on already overburdened government-provided insurance mechanisms and disaster assistance. Governments already are showing decreased willingness to assume new weather-related liabilities, and tensions concerning risk-sharing between local and federal government bodies also are evident (Fletcher, 2000)



Other reports in this collection