SDSS J084044.34+193743.9 , the SIMBAD biblio

X-ray emission from active galactic nuclei (AGN) is dominated by the accretion disk around a supermassive black hole. The radio luminosity, however, has not such a clear origin except in the most powerful sources where jets are evident. The origin (and even the very existence) of the local bi-modal distribution in radio-loudness is also a debated issue. By analysing X-ray, optical and radio properties of a large sample of type 1 AGN and quasars (QSOs) up to z>2, where the bulk of this population resides, we aim to explore the interplay between radio and X-ray emission in AGN, in order to further our knowledge on the origin of radio emission, and its relation to accretion. We analyse a large (∼800 sources) sample of type 1 AGN and QSOs selected from the 2XMMi XMM-Newton X-ray source catalogue, cross-correlated with the SDSS DR7 spectroscopic catalogue, covering a redshift range from z∼0.3 to z∼2.3. Supermassive black hole masses are estimated from the Mg ii emission line, bolometric luminosities from the X-ray data, and radio emission or upper limits from the FIRST catalogue. Most of the sources accrete close to the Eddington limit and the distribution in radio-loudness does not appear to have a bi-modal behaviour. We confirm that radio-loud AGN are also X-ray loud, with an X-ray-to-optical ratio up to twice that of radio-quiet objects, even excluding the most extreme strongly jetted sources. By analysing complementary radio-selected control samples, we find evidence that these conclusions are not an effect of the X-ray selection, but are likely a property of the dominant QSO population. Our findings are best interpreted in a context where radio emission in AGN, with the exception of a minority of beamed sources, arises from very close to the accretion disk and is therefore heavily linked to X-ray emission. We also speculate that the radio-loud/radio-quiet dichotomy might either be an evolutionary effect that developed well after the QSO peak epoch, or an effect of incompleteness in small samples.