We use recent progress in simulating the production of magnetohydrodynamic jets around black holes to derive the cosmic spin history of the most massive black holes .
Our work focusses on black holes with masses \mathrel { \hbox { \hbox to 0.0 pt { \hbox { \lower 2.0 pt \hbox { $ \sim$ } } } \raise 2.0 pt% \hbox { $ > $ } } } 10 ^ { 8 } M _ { \odot } .
Under the assumption that the efficiency of jet production is a function of spin \hat { a } , as given by the simulations , we can approximately reproduce the observed ‘ radio loudness ’ of quasars and the local radio luminosity function .

Using the X-ray luminosity function and the local mass function of supermassive black holes , SMBHs , we can reproduce the individual radio luminosity functions of radio sources showing high- and low-excitation narrow emission lines .
We find that the data favour spin distributions that are bimodal , with one component around spin zero and the other close to maximal spin .
The ‘ typical ’ spin is therefore really the expectation value , lying between the two peaks .
In the low-excitation galaxies , the two components have similar amplitudes , meaning approximately half of the sources have very high spins , and the other have very low spins .
For the high-excitation galaxies , the amplitude of the high-spin peak is typically much smaller than that of the low-spin peak , so that most of the sources have low spins .
However , a small population of near maximally spinning high-accretion rate objects is inferred .
A bimodality should be seen in the radio loudness of quasars , although there are a variety of physical and selection effects that may obfuscate this feature .

We predict that the low-excitation galaxies are dominated by SMBHs with masses \mathrel { \hbox { \hbox to 0.0 pt { \hbox { \lower 2.0 pt \hbox { $ \sim$ } } } \raise 2.0 pt% \hbox { $ > $ } } } 10 ^ { 8 } M _ { \odot } , down to radio luminosity densities \sim 10 ^ { 21 } W Hz ^ { -1 } sr ^ { -1 } at 1.4 GHz .
Under reasonable assumptions , our model is also able to predict the radio luminosity function at z = 1 , and predicts it to be dominated by radio sources with high-excitation narrow emission lines above luminosity densities \mathrel { \hbox { \hbox to 0.0 pt { \hbox { \lower 2.0 pt \hbox { $ \sim$ } } } \raise 2.0 pt% \hbox { $ > $ } } } 10 ^ { 26 } W Hz ^ { -1 } sr ^ { -1 } at 1.4 GHz , and this is in full agreement with the observations .

From our parametrisation of the spin distributions of the high- and low-accretion rate SMBHs , we derive an estimate of the spin history of SMBHs , which shows a weak evolution between z = 1 and 0 .
A larger fraction of low-redshift SMBHs have high spins compared to high-redshift SMBHs .
Using the best fitting jet efficiencies there is marginal evidence for evolution in spin : the mean spin increases slightly from \langle \hat { a } \rangle \sim 0.25 at z = 1 to \langle \hat { a } \rangle \sim 0.35 at z = 0 , and the fraction of SMBHs with \hat { a } \geq 0.5 increases from 0.16 \pm 0.03 at z = 1 to 0.24 \pm 0.09 at z = 0 .
Our inferred spin history of SMBHs is in excellent agreement with constraints from the mean radiative efficiency of quasars , as well as the results from recent simulations of growing SMBHs .
We discuss the implications in terms of accretion and SMBH mergers .
We also discuss other work related to the spin of SMBHs as well as work discussing the spin of galactic black holes and their jet powers .