Restrictions on managerial outside job opportunities and corporate tax policy: Evidence from a natural experiment