Tax credits for spending on research and development (R&D) made available by the Italian government after the 2008 credit crunch failed to stimulate any additional investment by high-tech firms. That is one of the findings of new research by Antonio Acconcia and Claudia Cantabene, which examines the impact of a short-term subsidy supplied to firms in bad times.
Their study, published in the August 2018 issue of the Economic Journal, shows that the effect of the policy was moderate on average and varied considerably across recipient firms, depending on their type – whether traditional or high-tech – and available liquidity.
Does countercyclical fiscal stimulus help to sustain demand? Despite its importance, the answer to this question remains highly controversial, mainly because of the difficulty in identifying the causal link empirically.
The new study contributes to the debate by looking at the case of a short-term subsidy supplied to firms in bad times. The researchers investigate the effect of a tax credit for R&D expenditure that the Italian government granted just after the 2008 credit crunch.
The main conclusion is that the effect of the policy was moderate on average and strongly heterogeneous among recipient firms, depending on their type – whether traditional or high-tech – and available liquidity.
The finding about heterogeneity accords with recent research findings on the effect of tax rebates among households. More generally, the evidence highlights the importance of the design of a policy for maximising its effect.
Two features qualify this case study as suitable for dealing with the identification issue. The first is related to the mechanism adopted to choose the subsidised firms. They were selected according to the chronological order in which the electronic applications of the candidates arrived at the website of the Italian Revenue Agency, given the fulfilment of an aggregate tax credit cap and without any project screening.
Since the aggregate constraint determined the rejection of many applications, the behaviour of those firms that applied for the stimulus but were not selected provides a useful frame of reference.
The second feature is the possibility for sorted firms to carry forward any amount of the credit in excess of current year tax to offset future taxes. By looking at recessionary years, the restriction to profitable firms would have raised the potential concern of a selection bias affecting the conclusion of the research.
Among high-tech firms, the 2008/09 change in the total amount of R&D expenditure by recipient firms was virtually the same as that of non-recipient firms, suggesting no effect for the public subsidy. By comparing firms of the traditional industries it follows, instead, that recipient firms increased expenditure in R&D by about 8-9%, while non-recipient ones did not change it.
Given this figure and considering that the tax credit determined a drop in the aggregate tax revenue as well as a 10% reduction of the R&D cost, and that high-tech firms at the time of the policy counted for less than 2% of the total number of firms in Italy, a simple back of the envelope calculation implies that by reducing the tax revenue by one euro the government determined more R&D by roughly the same amount.
It is arguably unpleasant that an economic stimulus for R&D did not drive larger expenditure among high-tech firms. But as the public policy studied here is a transitory subsidy, the unresponsiveness of such firms is consistent with theoretical prediction. Pushed by dynamic market competition, firms operating in high-tech industries tend to upgrade existing products and create new ones continuously. This suggests that there is a sluggish response of R&D to short-run changes in its cost.
The average effect of the tax credit estimated for the traditional firms masks large heterogeneity. Liquidity is the key dimension for identifying which firms are really affected by the economic stimulus. Recipient firms with relative large cash holdings raised their R&D expenditure quite a lot, even more if they were largely indebted at the time of the credit crunch. In contrast, recipients with small amounts of cash behaved as the corresponding group of non-recipient firms.
Figure 1 shows averages of R&D change when the sample of firms is split according to the median value of firms’ liquidity. Recipients with relatively high amounts of liquidity increased R&D – by about 0.6 percentage points of total assets. The R&D change of recipients with low liquidity as well as that of the non-recipient firms, whatever the amount of liquidity, is very close to zero.
Finally, it should be noted that when the tax credit encouraged more R&D, it induced a strong transitory effect on firm performance. This confirms that the average effect is indeed driven by a specific group of firms.
‘Liquidity and Firms’ Response to Fiscal Stimulus’ by Antonio Acconcia and Claudia Cantabene is published in the August 2018 issue of the Economic Journal.