We present the first population synthesis of synchronous magnetic cataclysmic variables, called polars, taking into account the effect of the white dwarf (WD) magnetic field on angular momentum loss. We implemented the reduced magnetic braking (MB) model proposed by Li, Wu & Wickramasinghe into the Binary Stellar Evolution (BSE) code recently calibrated for cataclysmic variable (CV) evolution. We then compared separately our predictions for polars and non-magnetic CVs with a large and homogeneous sample of observed CVs from the Sloan Digital Sky Survey. We found that the predicted orbital period distributions and space densities agree with the observations if period bouncers are excluded. For polars, we also find agreement between predicted and observed mass transfer rates, while the mass transfer rates of non-magnetic CVs with periods >~3h drastically disagree with those derived from observations. Our results provide strong evidence that the reduced MB model for the evolution of highly magnetized accreting WDs can explain the observed properties of polars. The remaining main issues in our understanding of CV evolution are the origin of the large number of highly magnetic WDs, the large scatter of the observed mass transfer rates for non-magnetic systems with periods >~3h, and the absence of period bouncers in observed samples.