This study provides new evidence on the factors associated with using
tax return preparers from a balanced panel of microlevel tax return da
ta for 1982-4. The results of fixed-effects logit models that control
for unobservable differences across individuals indicate that (1) the
time-cost of return preparation has a positive albeit small impact on
preparer use, (2) the type of tax return filed and employment status a
re important determinants of preparer use, and (3) the effect of incom
e and marginal tax rate on preparer use in models that control for com
plexity and source of income is unclear.