The idea that jobs could be created by an expansion of tax financed government expenditure can be extended beyond the realm of subsidies to low skilled private service activities. It is important to note that such a policy has different effects from those resulting from traditional deficit spending. The compensating rise in taxation chokes off the multiplier by redistributing spending power from existing taxpayers to those who find jobs as a result. e.g. in the public sector itself or within supplier sectors if the spending is on infrastructures. This limits the creation of jobs and, by the same token, leads only to a small rise in the demand for tradable;  however, it prevents any substantial worsening of the balance of payments the more so as the redistribution of consumption from higher income groups to the erstwhile unemployed should typically reduce import demand. The pattern of government spending can also be targeted to provide work for groups especially affected by the lack of jobs. Thus, the concentration of unemployment and non-employment on the less educationally qualified, or on those living in particular areas, can be taken into account in devising an appropriate pattern of expenditure increases. Within the public sector preference could be given to the long term unemployed, with confidence that this represented net additional jobs rather than the displacement of other workers. Obviously a balance has to be struck between the benefits of different spending patterns and their opportunity costs, but within public spending programmes the variation in social costs of employing different categories of people can be explicitly taken into account.