Factorization and Scaling in Hard Diffraction

We compare results on diffractive W-boson production at the Tevatron with predictions based on the diffractive structure function measured in deep inelastic scattering at HERA assuming (a) conventional factorization or (b) hard factorization combined with a rapidity gap distribution scaled to the total gap probability. We find that conventional factorization fails, while the scaling prediction agrees with the data.