Since they were first established in 1986, the national and provincial lists of poor counties have played an important role in organizing China's poverty reduction program and helped concentrate available poverty reduction funding in some of the areas of greatest need.

The 1993 revised list of 592 nationally-designated poor counties included an additional number of extremely poor counties which had not previously received an appropriate share of central government support.

However, the list of 592 poor counties is no longer an appropriate means of targeting available poverty reduction funding to China's remaining poor. "Targeting effectiveness has deteriorated over time and leakage has increased," and these trends were exacerbated by the new poor county　designations in 1993. It is now evident that the current system of county-based targeting results in (a) a severe dilution of available poverty reduction funding for the half of China's remaining poor

who reside in the nationally-designated poor counties, and (b) the near complete omission of　central government poverty reduction funding for the other half of the poor who reside outside　these poor counties. This severe problem could be largely resolved by switching to a system of　township-based targeting which would direct the bulk of available funding to poor townships　within and outside of the nationally-designated poor counties.

An ongoing SSB survey of China's nationally-designated 592 poor counties has shown that about half of China's remaining poor resided in these poor counties. This implies that in 1998 about 21 million of China's remaining 42 million poor were located in these counties.

Most of the central government's poverty reduction funding is allocated to the nationally-designated poor counties, but is not specifically earmarked for the poor households or the poor townships within these counties. Instead, these funds are distributed both to poor and non-poor townships within the poor counties. (Field visits suggest that non-poor townships and better off-villages closer in to the road system may in fact capture more than an equal share of the available poverty reduction funding.) Therefore, it appears that available poverty reduction funding is equally distributed to all 200 million rural inhabitants of the 592 poor counties. Since only about 21 million absolute poor resided in the poor counties in 1998, it is evident that assistance to the most needy was diluted by roughly ten-fold by the leakage of benefits to the non-poor. This suggests that in 1998 only about one tenth, or $200 million, of the US$2 billion of central government poverty reduction funding reached these 21 million poor.

Just as importantly, very little central government poverty reduction funding is provided　to the other half of the poor residing outside the poor counties (since most such funding is earmarked for use only in the nationally-designated poor counties). Instead, most provinces use　their own funding to assist the poor residing outside the poor counties. However, provincial　funding levels are known to be very limited. The problem is quite severe in Sichuan, and LGPR　has allowed the province to use 30 percent of its share of the central government's poverty reduction program funding to assist three extremely poor minority areas not included in the list of　nationally-designated poor counties. Overall, however, the total poverty reduction assistance　package available for the poor outside the nationally-designated poor counties is extremely　limited.

Targeting All Poor Townships. Most of China's remaining poor reside in poor townships within and outside of the nationally-designated poor counties. The Yunnan　PADO, for example, estimates that 86 percent of the province's remaining poor in 1996 were　concentrated in some 506 designated poor townships. (Some 396, or nearly four fifths, of these　designated poor townships are located in the province's 73 nationally- designated poor counties.) Therefore, as a most crucial first step, it is recommended that most central government poverty reduction funding should be channeled directly to poor townships within and outside of the 592　nationally-designated poor counties, and all poor villages should be guaranteed access to these　funds. Pilot tests of this switch from county to township based targeting should be undertaken　immediately, and need not wait until the completion of the 8-7 Plan" A switch to township.　targeting would　substantially reduce leakage to the non-poor, even in the absence of programs that more effectively target poor households.

Switching to township targeting would also benefit the minorities who presently live　outside the boundaries of both the poor counties and the minority regions, and thus currently　receive neither poverty assistance nor special types of relief administered in minority regions.

Concerns have been raised that a switch from county to township targeting would increase the administrative costs of the program. The most obvious implication of township targeting would be the need for additional staff at the township level. Lack of staff at this level is currently one of the major weaknesses of the poverty loan program, and switching to township targeting would certainly necessitate staffing at the township level. However, the costs of doing so appear modest, particularly when set against the improvements in targeting that are likely to occur with this switch.