2012MNRAS.423.2736N

The reported alignment between the projected spin-axes and proper motion directions of pulsars is revisited in the light of new data from Jodrell Bank and Effelsberg. The present investigation uses 54 pulsars, the largest to date sample of pulsars with proper motion and absolute polarization, to study this effect. Our study has found strong evidence for pulsar spin–velocity alignment, excluding that those two vectors are completely uncorrelated, with >99 per cent confidence. Although we cannot exclude the possibility of orthogonal spin–velocity configurations, comparison of the data with simulations shows that the scenario of aligned vectors is more likely than that of the orthogonal case. Moreover, we have determined the spread of velocities that a spin-aligned and spin-orthogonal distribution of kicks must have to produce the observed distribution of spin–velocity angle offsets. If the observed distribution of spin–velocity offset angles is the result of spin-aligned kicks, then we find that the distribution of kick-velocity directions must be broad with {img}; if the orthogonal-kick scenario is assumed, then the velocity distribution is much narrower with {img}. Finally, in contrast to previous studies, we have performed robustness tests on our data, in order to determine whether our conclusions are the result of a statistical and/or systematic bias. The conclusion of a correlation between the spin and velocity vectors is independent of a bias introduced by subsets in the total sample. Moreover, we estimate that the observed alignment is robust to within 10 per cent systematic uncertainties on the determination of the spin-axis direction from polarization data.