Power spectral analyses of the plasma density measured by the Voyager 2 spacecraft are used to investigate the spectral characteristics and fluctuation level of density turbulence from 1 to 60 AU, corresponding to the period 1977 to 1999. Only periods without clear discontinuities, with less than 10% missing data, and at least 256 samples are selected, strongly limiting the number and coverage of intervals analyzed. Typically, the power spectra have a dual power law form, with a closely Kolmogorov component at higher frequencies and a steeper component ($propto$ f-2) at lower frequencies, consistent with observations inside 1 AU. The spectral indices remain approximately constant with heliocentric distance R $lesssim$ 30 AU, with weak evidence for a decrease toward the Kraichnan index at larger R for the h component. For R $lesssim$ 20 AU the fluctuation level (δn)2(R) decreases approximately as R-3¿1 but then flattens out in the region 20--30 AU before increasing slowly for R $gtrsim$ 30 AU. This injection of energy into density turbulence beyond 20--30 AU is not a heliolatitude effect and is contrary to the R-4 behavior for (δn)2(R) predicted on the basis of observations inside 1 AU. It is interpreted as the result of charge-exchange and pickup of interstellar neutral atoms for R > 1 AU driving enhanced MHD turbulence and associated density variations. The results provide independent support that observations of proton heating and enhanced magnetic turbulence levels beyond about 20 AU are both real and associated with pickup ions. They also suggest that existing calculations of angular scattering of the 2--3 kHz radiation observed by the Voyagers in the outer heliosphere must be redone and provide opportunities to test turbulence theories by comparing density and magnetic turbulence as a function of R in the outer heliosphere. Owing to limitations on the coverage and quality of data, confirmation of these results using nonspectral methods is desirable.