Should there be tax relief/gift aid for charitable donations? Does it even promote more giving? This is not known.

Key points from my response to the Charity Tax Commission call for evidence

Presumably the main argument for offering tax relief and Gift Aid is that this promotes more giving, i.e., convinces people to donate more from their own pockets, to take advantage of these benefits.

Much economic research in fact considers how the amount people choose to give responds to the so-called ‘tax-price’  of giving (the amount of after-tax income an individual must give up in order for the charity to gain £1). If this elasticity is below the ‘gold standard’  this implies that by offering incentives (such as Gift Aid and higher rate relief), the government is actually reducing the amount that people choose to donate out of their own pockets. In such a case it would be more efficient to eliminate Gift Aid etc., and have the government  donate this money directly to good causes.

As an aside,  this would raise the question, ‘which causes should be chosen for support?’ The government might decide to support the charities that people donate to, perhaps  proportionally to the amounts donated. In theory (ignoring the issue of higher rate taxpayers) this latter policy would actually give people the same incentives as above, so it would effectively not change anything.

Evidence on this elasticity

While a great deal of work has been done, ultimately the evidence is mixed as to whether the elasticity in fact exceeds this gold standard. It is difficult to identify the response to the “tax price” of giving for a variety of reasons.  In particular, there have been few if any experiments in altering the rebate/match at a national level that would permit such clean evidence.

There is some evidence from particular fundraising appeals; from non-representative circumstances. This offers limited evidence that the presence of a match increases individualsout-of-pocket’ donations. </span><a href="http://f1000.com/work/citation?ids=2943034&amp;pre=&amp;suf=&amp;sa=0"><span style="font-weight:400;">Karlan and List (2007)</span></a><span style="font-weight:400;">, find evidence that it only increases this for one subset of respondents, and they dont seem to consider the actual marginal impact of the match rate; the evidence suggests that any effect is being driven by the psychological impact of the presence of an outside supporter. Other such work finds it may even be counterproductive (Huck and Rasul 2011); and it would work better better to be simply a ‘lead gift’. Meier (2007) finds that the positive (participation) effect may be crowded-out in the longer run when the match is removed.

Other evidence (e.g., Feldstein and Clotfelter 1976; Auten, Sieg, and Clotfelter 2002; Backus and Grant 2018; Hungerman and Ottoni-Wilhelm 2016 survey this) comes from naturally occurring observational data on people’s tax rates and level of giving, and usually relies on particular parts of tax data where giving is ‘declared’ (in the US, those who ‘itemise their deductions’). The latter tends to rely on natural variations across US states and over time, in response to changes in legislation, etc. However, my impression is that this has rarely been done an extremely convincing way. A major ‘econometric identification’ issue is that it is difficult to disentangle a nonlinear income elasticity of giving from a response to a change in the tax price.