Detroit effort puts priority on ending homelessness...

[emu steve]

Sticking with poverty for now, I found this argument of decreasing poverty surprising. Quick research suggests that the poverty rate in the 1950 thru 1960 was dropping like a rock, and then has levelled off since then. So I'm not sure the statistics are valuable in assessing anything in the last, say, 40 years.

I'm also curious how the stats look for urban poverty. A lot of the poverty in the US in the 50s and 60s was rural poverty. A lot was white. Much Appalachian.

Anyone got any real stats? I had a hard time getting past the vast volume of right-wing stats saying thing suck. And some left-wing stats saying things are great. Census bureau anyone?

It's a shame but the Census Bureau [[I used to work there) produced something called "Social Indicators" in the 70s and 80s. My friend, a demographer, wrote several of the chapters, for Social Indicators III. [I used to own Social Indicators II but, of course, don't have it now.]

It had the best longitudinal data on about 11 different areas such as employment, educational attainment, health, income, crime, leisure and recreation, etc. etc.

It got canned when the Census Bureau budget took a real budget whacking in 1982 [[lot of employees lost their jobs - yes, federal employees have lost their jobs in budget shortfalls).

It was great for bring together national longitudinal data for a lot of much debated topics. Lot of great charts.

For those not familiar with the Census Bureau [[other then the decennial census), they have a very extensive data collection operation. When I was there they had a division [[approximately 175 headquarters employees) working on demographic surveys [[usually annual but some monthly) for BLS [[for the unemployment numbers. That is where the 5.4% unemployment number comes), Housing for HUD, health for HHS, crime for DOJ, consumer expenditures for BLS, etc. etc.

Unfortunately, it isn't really used well today in public policy debates. The data are still there...

Getting back to crime. Census did a 'criminal victimization survey' for DOJ. The most accurate crime data is homicide. That is very accurate data. The next best is probably this survey where respondents are actually asked if they have been a victim of crime. These data are probably much better than the FBI uniform crime report data. After all who know better if they were assaulted, robbed, had their house broken into, etc. then the victim???