We discuss the nuclear EMC effect with particular emphasis on recent data for light nuclei including $^{2}\mathrm{H}$, $^{3}\mathrm{He}$, $^{4}\mathrm{He}$, $^{9}\mathrm{Be}$, $^{12}\mathrm{C}$, and $^{14}\mathrm{N}$. In order to verify the consistency of available data, we calculate the ${\ensuremath{\chi}}^{2}$ deviation between different data sets. We find a good agreement between the results from the NMC, SLAC E139, and HERMES experiments. However, our analysis indicates an overall normalization offset of about $2%$ in the data from the recent JLab E03-103 experiment with respect to previous data for nuclei heavier than $^{3}\mathrm{He}$. We also discuss the extraction of the neutron/proton structure function ratio ${F}_{2}^{n}/{F}_{2}^{p}$ from the nuclear ratios $^{3}\mathrm{He}$/$^{2}\mathrm{H}$ and $^{2}\mathrm{H}$/$^{1}\mathrm{H}$. Our analysis shows that the E03-103 data on $^{3}\mathrm{He}$/$^{2}\mathrm{H}$ require a renormalization of about $3%$ in order to be consistent with the ${F}_{2}^{n}/{F}_{2}^{p}$ ratio obtained from the NMC experiment. After such a renormalization, the $^{3}\mathrm{He}$ data from the E03-103 and HERMES experiments are in a good agreement. Finally, we present a detailed comparison between data and model calculations, which include a description of the nuclear binding, Fermi motion, and off-shell corrections to the structure functions of bound proton and neutron, as well as the nuclear pion and shadowing corrections. Overall, a good agreement with the available data for all nuclei is obtained.