Investor-Level Taxes and Corporate Dividend Policy: Evidence from a Quasi-Natural Experiment

We combine a change to the treatment of franking credits in the hands of
domestic shareholders, namely the introduction of fully refundable franking
credits, and a novel methodological approach to provide robust evidence of the
causal effect of investor-level taxes on corporate dividend policy. Consistent
with investors having a greater preference for the distribution of dividends, we
find that the introduction of fully refundable franking credits increased both
the likelihood that firms paid dividends and the level of the di vidend payments
they made. Subsequent analysis reveals that the effective tax rate and firm size
explain cross-sectional variation in dividend policy responses to the tax reform.
Specifically, testing shows that large firms and firms with high effective tax rates
were more likely to pay dividends, and to pay larger dividends, than their peers in
response to the tax reform.