^^ Mr. Paral's estimates are not just predictions; they are simply based on the Census Bureau's estimates of population, which are well-documented and published. We've always known that the Census Bureau tends to undercount urban populations in their estimates; more accurate numbers only come around every 10 years during actual censuses. Many cities routinely challenge the Census Bureau's population estimates, but AFAIK, Chicago has not challenged since 1990.
It's actually interesting, because
if you look at Paral's survey (PDF), the data implies that Chicago lost 5% of its population between 2000 and 2006. So if Mr. Paral is predicting a 5% drop between 2000 and 2010, that means that Chicago's population held steady in 2006-2010, which is definitely a good sign.
I didn't know that we gained population in the 90s, though. That's interesting; I always thought Chicago was on a continual downward trend that we had yet to reverse. The
University of Chicago claims that the 90s gains were largely due to Hispanic immigration and the repopulation of industrial or blighted areas by the professional class, and were mitigated by the flow of blacks out of the city. So a reversion back to population loss either implies that immigration was reduced (Bush crackdowns on illegals?), or that blacks started leaving in much higher numbers than they did in the 90s (00's crime wave?)
The growth of wealthy areas on the north/northwest sides and around the Loop was going gangbusters in the 00's, so that can't be the reason for population loss, unless the smaller household sizes of the new wealthier residents combined with their displacement of the former residents was a substantial factor. Personally, I see taller buildings all over the North Side, and it seems like the greater number of units in the new, taller buildings should offset any drop in household size, for no significant change in density in neighborhoods pre- and post-gentrification.