Cosmic fluctuation spectra with large-scale power

Observations of the cluster-cluster and cluster-galaxy correlation functions and of large-scale streaming velocities seem to indicate that more power exists on large scales than is predicted in the standard Omega = 1 cold dark matter (CDM) model with scale-invariant initial conditions. Although these observations are difficult and the results are not firmly established, these tests can in principle be used to discriminate sharply among cosmological models. The extra power required to explain the reported observations is explored by systematically varying cosmological models having initially Gaussian perturbations. Low Hubble constant models, hybrid models with hot and cold dark matter, models with both adiabatic and isocurvature modes present, and decaying dark matter models are shown to fail these observational tests. Models that are open or have a large cosmological constant do better, but still fail. A CDM model with a short-lived massive neutrino does have enough large-scale structure, although the origin of the required decay time is quite speculative. Microwave background anisotropies at a few degrees must be on the verge of detection if the required extra power exists.