We have made radio observations of 87 optically selected quasars at 5 GHz with the VLA in order to measure the radio power for these objects and hence determine how the fraction of radio-loud quasars varies with redshift and optical luminosity. The sample has been selected from the recently completed Edinburgh Quasar Survey and covers a redshift range of 0.3 < z < 1.5 and an optical absolute magnitude range of -26.5 < M_{B} < -23.5 (h, q_{0} = 1/2). We have also matched up other existing surveys with the FIRST and NVSS radio catalogues and combined these data so that the optical luminosity-redshift plane is now far better sampled than previously. We have fitted a model to the probability of a quasar being radio-loud as a function of absolute magnitude and redshift and from this model infer the radio-loud and radio-quiet optical luminosity functions. The radio-loud optical luminosity function is featureless and flatter than the radio-quiet one. It evolves at a marginally slower rate if quasars evolve by density evolution, but the difference in the rate of evolutions of the two different classes is much less than was previously thought. We show, using Monte-Carlo simulations, that the observed difference in the shape of the optical luminosity functions can be partly accounted for by Doppler boosting of the optical continuum of the radio-loud quasars and explain how this can be tested in the future.