Lets leave aside for a moment what the levels of social expenditure might be. The composition of these also matters I contend. At one extreme we can give people a lot of cash and allow them to go find the services they need. At another extreme we can give them very little cash and provide the services. This debate comes into focus now and again when people in Ireland compare welfare rates here with those in Northern Ireland. The argument is usually “here if you are in receipt of [insert welfare program name] you get X and in the north you would only get Y (Y << X), its a disgrace, we should slash [insert welfare program name]”. This of course ignores that the cash benefit versus the in kind benefit makeup of the comparators might well and usually does differ
Below is a chart taken from OECD data on this composition. A high level means lots of benefits in kind and not much cash (as a percentage of total expenditure). Its for all programmes, and show this issue in fairly stark terms. Note how things have changed in Ireland over the years. We show the biggest variation over time of all. Initially we gave people things, then we gave cash then as we got rich we gave things now we are back to cash. This is hardly indicative or supportive of a consistent and coherent policy on social expenditure.