Implications of Precision Electroweak Data

There are two aspects to the 1995 summer update of the combined preliminary electroweak data from LEP and SLC. On the one hand, agreement between experiments and the Standard Model (SM) has improved for the line-shape and the asymmetry data. The $\tau$ widths and asymmetries are now consistent with $e$--$\mu$--$\tau$ universality, and all the asymmetry data including the left-right asymmetry from SLC are consistent with the SM (16\%CL). On the other hand, a discrepancy between experiments and SM predictions is sharpened for two observables, $R_b$ and $R_c$, where $R_q$ is the partial $Z$ boson width ratio $\Gamma_q/\Gamma_h$. $R_b$ is 3\% larger (3.7$\sigma$) and $R_c$ is 11\% smaller (2.5$\sigma$) than the SM predictions. When combined, the SM is ruled out at the 99.99\%CL for $m_t>170$GeV. It is difficult to interpret the $11\pm 4$\% deficit of $R_c$, since if we allow only $\Gamma_b$ and $\Gamma_c$ to deviate from the SM then the precisely measured ratio $R_h=\Gamma_h/\Gamma_\ell$ forces the QCD coupling to be $\alpha_s\equiv \alpha_s(m_Z)_{\overline{\rm MS}}=0.185\pm 0.041$, which is uncomfortably large. The data can be consistent with the prefered $\alpha_s$ ($0.10<\alpha_s<0.13$) only if the sum $\Gamma_h=\Sigma_q \Gamma_q$ does not deviate significantly from the SM prediction. Possible experimental causes for the under-estimation of $R_c$ are discussed. By assuming the SM value for $R_c$, the discrepancy in $R_b$ decreases to 2\% (3$\sigma$). The double tagging technique used for the $R_b$ measurements is critically reviewed. A few theoretical models that can explain large $R_b$ and small $\alpha_s$ ($= 0.104\pm 0.008$) are discussed. If the QCD coupling $\alpha_s$ is allowed to be fitted by the data, the standard $S$, $T$, $U$ analysis for a new physics