Suppose that you were a member of a chama with 10 members. I tell you that I want to support your chama by giving each member Sh1000. In addition, I will give the chama as a whole Sh10,000 to share equally among the members. On top of that, I want to give each member who has children Sh1000, and each member who does not have children SH1000.
You might be happy to take my money, but you might also wonder why I have made things so complicated. In the end, I am just giving each member of your chama Sh3000 (Sh1,000 directly, another Sh1,000 after the Sh10,000 is divided up, and then Sh1,000 to each member with and without children, which is all of you).
This might seem an unlikely scenario, but when governments start to give out money without an overall policy, they can end up creating a complex set of grants that are difficult to administer and may not achieve what they intended. If I give people with no children a grant to control population growth, and people with children a grant to support their school fees, in the end I am not incentivizing anything. People with and without children are both equally better off after my grant than before I started handing out the cash. I have also wasted resources creating a system to monitor whether people have children (to know which grant they qualify for), even though there is no impact of differentiating between them. In the end, from the perspective of changing behavior, this situation is worse than not giving any grants at all.
As the government advances its policy agenda and implements devolution, it is already relying on a series of conditional transfers without any framework. The 2010 constitution recognizes conditional grants to counties, but good financial practice is to consider the total resources available to achieve policy goals and then decide how this money should be shared. Some of it is already going to counties as part of their “equitable share,” the constitutionally mandated 15% minimum, which is an unconditional grant (meaning counties may spend it anyway they like).
The key question is then how to augment what the constitution requires with additional funds that achieve something beyond what the equitable share can achieve. The basic principle of using different tools for sharing money is that we are trying to achieve different things. As I described above, I might want to encourage people to have fewer children. Or I might want to support families with children to ensure their children have access to schooling. In designing a conditional transfer, I must clearly identify the purpose of the transfer and explain how it relates to other transfers, else it might not achieve its aim. I can give people money to have fewer children, or to support their school fees, but doing both largely cancels out. This is why a framework is needed.
So far, Kenya’s record on the use of conditional transfers falls short. The government appears to have accepted the recommendation of the Commission on Revenue Allocation to use the Equalisation Fund as a conditional grant to 14 “marginalized” counties. The effect of this is largely to favor the same counties that are already favored by the formula. Why use a conditional grant to do what your unconditional transfer already does? This violates the principle of using different tools to do different things.
Then there is the Constituency Development Fund (CDF). Last week, the dailies noted without comment that CDF money is to be distributed according to the same formula which determines the distribution of revenues to counties. In other words, yet another conditional grant is being used to favor the same places that benefit from the unconditional transfer. True, there are some differences because the formula is applied to the constituency rather than the county level, but there is no analysis to suggest that this is the best way to deal with the issues that the unconditional transfer is unable to address.
What issues might those be? One is that there are huge inequalities within counties. For example, access to improved water is an area of stark inequality within all 47 counties. Moreover, water access inequality is more severe across wards than constituencies. It might be best handled by a conditional transfer to counties to be spent on the wards that are most deprived. The money might be targeted to the counties with the highest number of deprived wards, rather than the worst overall access at county level. A water grant like this would be consistent with our motto: use different tools to solve different problems.
This is just an example. The point is: it is past time to start talking about how to design and use conditional transfers.
Posts
No comments:
Post a Comment