With shrinking resources and declining federal transfers, provincial governments across Canada are forced to provide increased levels of supports to vulnerable individuals with decreasing resources (Janssen & Estevez, 2013). Governments continue to face obstacles in meeting the needs of vulnerable populations such as children, single parents, and those who are homeless, to name a few. Manitoba, for instance, faces demographic challenges related to an influx of newcomers who are seeking refuge, resettlement, and housing supports, an aging baby boomer population that will need end-of-life supports, as well as a growing number of children in government care. Instead of funding programs based on their activities and outcomes, this paper presents outcomes-based financing, such as the social impact bond, that reward service providers who are able to demonstrate proof of outcomes and can show how the intervention improved the lives of the individuals it was meant to serve. Under a social impact bond, government engages non-traditional partners in the private and non-for-profit sectors, and the community as a whole becomes part of the solution to challenging social problems.

A factorized finite-difference scheme for numerical approximation of initial-boundary value problem for two-dimensional subdiffusion equation in nonhomogeneous media is proposed. Its stability and convergence are investigated. The corresponding error bounds are obtained.

-

We consider Poisson's equation on the unit square with a nonlocal boundary condition. The existence and uniqueness of its weak solution in Sobolev space H 1 is proved. A finite difference scheme approximating this problem is proposed. An error estimate compatible with the smoothness of input data in discrete H 1 Sobolev norm is obtained.

A factorized finite-difference scheme for numerical approximation of initial-boundary value problem for two-dimensional fractional in time diffusion equation is proposed. Its stability is investigated and a convergence rate estimate is obtained.