Evidence that local tax and expenditure limits (TELs) for public K-12 schools lower student achievement is widely attributed to the effects of reduced funding, but our results cast doubt on reduced funding as the primary explanation for negative effects of TELs and instead suggest the importance of the predictability of funding. Students in districts subject to more severe local tax limits in Oregon score less well on eighth-grade tests in mathematics, but reduced funding is not the reason. Our analysis expands prior work by accounting for the extent to which TELs are actually binding, as well as for both pecuniary and nonpecuniary effects of TELs. Distinguishing pecuniary and non-pecuniary effects allows us to document that the negative effect of TELs in Oregon is not due to reduced expenditures. The state's school-finance equalization (SFE) tends to offset funding differentials, so TELs have no significant effect on funding, but even if TELs did affect funding, the negative effect of TELs on achievement is significant even if district expenditures are held constant. Instead, the negative effect of more restrictive TELs appears to work by disrupting local planning. We isolate this effect by distinguishing the more uncertain first year of each biennial budget from the second year. Our quasi-experimental design accounts for district and year fixed effects, as well as for district-specific variations in expenditures and student attributes. Results are robust to a placebo test designed to reveal spurious correlation and to several alternative specifications.