We provide evidence on the effect of property transaction taxes (stamp duty) on house prices, the timing of house purchases, and the volume of house purchases. To address these questions, we exploit administrative stamp duty records covering the universe of property transactions in the UK from 2004-2012 along with compelling sources of quasi-experimental variation. First, discontinuous jumps in the stamp duty at threshold property prices -notches- allow us to estimate the effect of the tax on house prices. Second, anticipated and unanticipated changes in the tax schedule allow us to estimate the dynamics of price responses and timing effects on house transactions. Third, a stimulus program that temporarily exempted certain properties from tax -a stamp duty holiday- allows us to provide micro evidence on both timing and extensive margin responses to macro stimulus policy. We find that the effect of transaction taxes on house prices is large (200-500% of the tax itself) and that adjustment to changes in transaction taxes is very fast. We also find that the timing of house transactions responds sharply to anticipated tax increases. Finally, temporary cuts in transaction taxes successfully stimulate housing market activity in the short run -an elimination of the tax increases transaction volume by about 20%- followed by a slump in activity after the policy is withdrawn. However, post-stimulus reversal is far from complete (30-40% of the initial boost), implying that stimulus has a long-run extensive margin effect in addition to a pure timing effect. This contradicts recent findings in the literature and has potentially important
implications for evaluating macro stimulus policy. (Joint work with Michael Carlos Best.)