Perturbative contributions to $\Delta\alpha^{(5)}(M^2_Z)$

We compute a theoretically driven prediction for the hadronic contribution to the electromagnetic running coupling at the $Z$ scale using lattice QCD and state-of-the-art perturbative QCD. We obtain$$\Delta\alpha^{(5)}(M^2_Z)=\left[279.5\pm0.9\pm0.59\right]\times10^{-4}\quad\quad\,\,\,\,\,\,(\mathrm{Mainz \,\,\,Collaboration})$$$$\Delta\alpha^{(5)}(M^2_Z)=\left[278.42\pm0.22\pm0.59\right]\times10^{-4}\,\,\,\,\,\,\,\,\quad(\mathrm{ BMW \,\,\,Collaboration}),$$ where the first error is the quoted lattice uncertainty. The second is due to perturbative QCD, and is dominated by the parametric uncertainty on $\hat{\alpha}_s$, which is based on a rather conservative error. Using instead the PDG average, we find a total error on $\Delta\alpha^{(5)}(M^2_Z)$ of $0.4\times10^{-4}$. Furthermore, with a particular emphasis on the charm quark contributions, we also update $\Delta\alpha^{(5)}(M^2_Z)$ when low-energy cross-section data is used as an input, obtaining $\Delta\alpha^{(5)}(M^2_Z) = \left[276.29 \pm 0.38 \pm 0.62\right] \times 10^{-4}$. The difference between lattice QCD and cross-section-driven results reflects the known tension between both methods in the computation of the anomalous magnetic moment of the muon. Our results are expressed in a way that will allow straightforward modifications and an easy implementation in electroweak global fits.