EclectEcon

A recent study for the C.D. Howe Institute is called, "The Welfare Enigma: Explaining the Dramatic Decline in Canadians’ Use of Social Assistance, 1993–2005," by Ross Finnie and Ian Irvine. The title is misleading, though, for there is no enigma: welfare benefits were reduced, employment options improved, and the combination meant fewer people sought social assistance.

...the SA rate fell, from a peak of 3.1 million individuals
in the early 1990s to 1.7 million in 2005.

In other words, as social assistance benefits fell, and as the opportunity costs of going on social assistance rose [eco-speak for saying that people had improved options, compared with going on welfare], surprise! Fewer people chose to go on social assistance.

The importance of these empirical findings is to see that welfare is not an either-or thing; rather, adjusting the height of the social safety net plays a role in determining how many will avail themselves of the support provided by that net. And if we opt for a lower social safety net, fewer people will use it.

For further evidence along these same lines, see this by Tim Worstall at the Adam Smith Institute.