U.S. CitiesCity-Data Forum Index
Go Back   City-Data Forum > General Forums > Urban Planning
 [Register]
Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
View detailed profile (Advanced) or search
site with Google Custom Search

Search Forums  (Advanced)
Reply Start New Thread
 
Old 02-05-2015, 02:59 PM
 
2,748 posts, read 2,697,239 times
Reputation: 3011

Advertisements

Gentrification in America Report

Some interesting insight and information about genrification in the US from Governing magazine that was just released
Reply With Quote Quick reply to this message

 
Old 02-06-2015, 01:24 AM
 
2,388 posts, read 2,954,341 times
Reputation: 1953
Nice post.

Quote:
Neighborhoods gentrifying since 2000 recorded population increases and became whiter, with the share of non-Hispanic white residents increasing an average of 4.3 percentage points. Meanwhile, lower-income neighborhoods that failed to gentrify experienced slight population losses and saw the concentration of minorities increase. They have also experienced different economic fates: Average poverty rates climbed nearly 7 percent in already lower-income tracts that didn’t gentrify, while dropping slightly in gentrifying neighborhoods.
A 4% increase in white people and <1% drop in the poverty rate

ohmygod!!!! displacement!!!!
Reply With Quote Quick reply to this message
 
Old 02-06-2015, 08:46 AM
 
56,593 posts, read 80,890,793 times
Reputation: 12500
It also mentioned that gentrification occurs in select cities in certain regions and using the 50 biggest cities in the country will include Sun Belt cities and even suburbs.
Reply With Quote Quick reply to this message
 
Old 02-06-2015, 11:34 PM
 
1,478 posts, read 2,002,048 times
Reputation: 1579
Quote:
Originally Posted by drive carephilly View Post
Nice post.



A 4% increase in white people and <1% drop in the poverty rate

ohmygod!!!! displacement!!!!
Keep in mind poverty is about 20% more frequent today than it was in 2000.

I was talking to a spacial demographer about poverty estimates. I don't want to butcher his words here, but the field is having a really difficult time swallowing some of the ACS data, particularly as it relates to income/poverty. The margins of error are huge. Just as an example, using the 5-year data the authors used in the report by family income group for my tract (which is gentrifying):

$10K to 15K: 6.0% of total families. MOE = +/- 6.1%
15-25: 13.1%, +/- 9%
25K-35K: 5.7% +/-4.2%

When your MOE is the majority of (or in some cases even greater) than your estimate, your estimates are pretty iffy.

The samples suck because a) the areas are small b) they can't sample enough at any point in time (relying upon a 5 year average) c) there is a "drift" since the 2010 census gave them a population base to work from, ie, there is an extra degree of freedom because they're basing a guesstimate of income stratification on a guesstimate of total people.

There has been some talk of figuring out a better way to get this data. Should the census expand the sample to get meaningful data on a regular basis (the goal of the ACS)? Should we just have a bigger sample from a long form census to measure during traditional census years? Should we do the long form/short form census split in traditional years and then do a more comprehensive study in years ending in 5 midway through?

FWIW, I do think the displacement argument is overblown outside of maybe a half dozen or so really large/high income metros. If someone gets pushed 2 miles down the road in a "normal" metro, it's not a economic hardship the vast majority of the time.
Reply With Quote Quick reply to this message
 
Old 02-07-2015, 01:32 AM
 
2,388 posts, read 2,954,341 times
Reputation: 1953
Quote:
Originally Posted by Chicago76 View Post
Keep in mind poverty is about 20% more frequent today than it was in 2000.
Sure. But if you lost your shirt in the recession, even if you weren't doing well to begin with, the chances are you were moving whether the rent goes up or not.


Quote:
The samples suck because a) the areas are small b) they can't sample enough at any point in time (relying upon a 5 year average) c) there is a "drift" since the 2010 census gave them a population base to work from, ie, there is an extra degree of freedom because they're basing a guesstimate of income stratification on a guesstimate of total people.

There has been some talk of figuring out a better way to get this data. Should the census expand the sample to get meaningful data on a regular basis (the goal of the ACS)? Should we just have a bigger sample from a long form census to measure during traditional census years? Should we do the long form/short form census split in traditional years and then do a more comprehensive study in years ending in 5 midway through?
I started to notice that the quality of the data sucked in '02 when the 2000 data started to come in and was comparing it to the '98 estimates. I think the answer is a 5 year census. It's ridiculous that 30 million people arrive in this country and we miss it because we wait a decade to do a census.
Reply With Quote Quick reply to this message
 
Old 02-08-2015, 09:03 PM
 
Location: Vallejo
14,061 posts, read 16,078,369 times
Reputation: 12636
Quote:
Originally Posted by drive carephilly View Post
I started to notice that the quality of the data sucked in '02 when the 2000 data started to come in and was comparing it to the '98 estimates. I think the answer is a 5 year census. It's ridiculous that 30 million people arrive in this country and we miss it because we wait a decade to do a census.
Nobody misses it. There's still estimates. The 1999 estimate was 272.5 million increasing by roughly 2.5 million per year. 2000 Census was 281 million, above what you'd expect the estimate of about 275 million definitely. Far cry from this 30 million, however.

The census is expensive. It cost $42 per person to do the 2010 census. Obviously having a yearly census from a data perspective would be nice. The question is is it worth it. I think ten years is about right given the cost.
Reply With Quote Quick reply to this message
 
Old 02-09-2015, 10:35 AM
 
1,478 posts, read 2,002,048 times
Reputation: 1579
Quote:
Originally Posted by drive carephilly View Post
Sure. But if you lost your shirt in the recession, even if you weren't doing well to begin with, the chances are you were moving whether the rent goes up or not.
This is the subtle distinction in the displacement argument that many ignore (and I agree with). Poor households tend not to stay in one place very long regardless, so displacement in many cases would have occurred regardless, even in a good economy. Displacement a mile down the road in a run of the mill large metro isn't particularly damaging. The ugly side of displacement more frequently occurs in expensive metros: someone gets booted from SF city because the affordable housing provision isn't there any more and ends up living miles away because there is no substitute. If they can't find something else in SF proper or in central Oakland (which is pretty in demand as far as affordable housing options go), then commutes are a nightmare.

Still, I wouldn't draw too many conclusions about displacement from that almost zero decline in poor households. Poor census tracts contain a lot of households somewhere between poverty status and living wage status. These households the ones generally falling back into poverty. They muddle the modest decline in poor household frequency in gentrifying census tracts substantially. The data point is essentially worthless. You'd really need a longitudinal study of households in both gentrifying and non-gentrifying census tracts to determine the impact.



Quote:
Originally Posted by Malloric View Post
Nobody misses it. There's still estimates. The 1999 estimate was 272.5 million increasing by roughly 2.5 million per year. 2000 Census was 281 million, above what you'd expect the estimate of about 275 million definitely. Far cry from this 30 million, however.

The census is expensive. It cost $42 per person to do the 2010 census. Obviously having a yearly census from a data perspective would be nice. The question is is it worth it. I think ten years is about right given the cost.
Aggregate numbers of basically everyone are much easier to estimate fairly closely. 2-3% error for the entire population is okay. Where you introduce large errors is when looking at smaller geographic areas or subsets of the population. Where errors get really large is in small subsets of small areas. Getting these areas right actually carries a lot of weight. Economic development funds/grants, school funding, etc is governed by income and population issues at the local level. That's why it is such a big deal. We are relying upon figures that may state for a given census tract that the poverty rate is 20% +/-12%. Relying upon a survey that estimates an area to have a poverty rate of 8% to 32% is worthless when it comes to funding decisions. The MOE is primarily a function of sample size. We hit 1 in 6 households with the long form in 2000. We now hit only 2% of households with the ACS every year.

Running a standard census that doesn't contain long form questions every 5 years isn't really going to help much. Bringing back the long form for 2020 and then doing another long form only update in 2025 would definitely help though. The other issue of course is expense. The census needs to do a better job of finding more cost effective information gathering methods. Electronic submission of household data was not available in 2010. That is absolutely ridiculous. This would free up field survey teams to hit lower income areas more heavily and it would reduce postage costs. Tying census completion to something to other registration services not explicitly protected (aka, voting) would also help. Need a DL renewed? The branch could enter your name to make sure you have been included in a household whose forms have been submitted. Otherwise, you need to get your forms in before getting the DL. These are just little things that would help.
Reply With Quote Quick reply to this message
 
Old 02-10-2015, 01:45 PM
 
2,388 posts, read 2,954,341 times
Reputation: 1953
Quote:
Originally Posted by Chicago76 View Post
This is the subtle distinction in the displacement argument that many ignore (and I agree with). Poor households tend not to stay in one place very long regardless, so displacement in many cases would have occurred regardless, even in a good economy. Displacement a mile down the road in a run of the mill large metro isn't particularly damaging. The ugly side of displacement more frequently occurs in expensive metros: someone gets booted from SF city because the affordable housing provision isn't there any more and ends up living miles away because there is no substitute. If they can't find something else in SF proper or in central Oakland (which is pretty in demand as far as affordable housing options go), then commutes are a nightmare.
True. I've made this point a few times in different threads. That this sort of thing is relatively rare and mostly happening in the most expensive cities with a higher than normal % of renters. SF is a pretty extreme example because of the geography.

It's clear that things are changing in a lot of cities but, for the most part, the changes a lot more subtle than they're made out to be and in a lot of cases critics are pointing to rapid changes in demographics brought on by infill development as evidence of displacement. When you have 700 middle class white people show up in a new apartment complex built on a former parking lot in a neighborhood that used to be home to 500 people it changes the stats quite a bit.

Quote:
Still, I wouldn't draw too many conclusions about displacement from that almost zero decline in poor households. Poor census tracts contain a lot of households somewhere between poverty status and living wage status. These households the ones generally falling back into poverty. They muddle the modest decline in poor household frequency in gentrifying census tracts substantially. The data point is essentially worthless. You'd really need a longitudinal study of households in both gentrifying and non-gentrifying census tracts to determine the impact.
I've actually done quite a bit of that going back to the 1940 census. Without interviewing people you can never know exactly why people moved but the data will make it obvious that certain things are or aren't happening.

I'm not really drawing conclusions from this single data point. I'm looking at the two together - 4% is kind of a joke in the context of the hair-on-fire displacement argument. Especially when you're looking at very modest declines in poverty. I recognize that the last decade has been a bad one for poverty but I'm also putting this in the context of all of the other data I've seen.

How Governing got it wrong: The problem with confusing gentrification and displacement | City Observatory


Quote:
Electronic submission of household data was not available in 2010. That is absolutely ridiculous. This would free up field survey teams to hit lower income areas more heavily and it would reduce postage costs. Tying census completion to something to other registration services not explicitly protected (aka, voting) would also help. Need a DL renewed? The branch could enter your name to make sure you have been included in a household whose forms have been submitted. Otherwise, you need to get your forms in before getting the DL. These are just little things that would help.
Boom.
Reply With Quote Quick reply to this message
 
Old 02-12-2015, 09:11 AM
 
33,046 posts, read 22,053,448 times
Reputation: 8970
Quote:
Originally Posted by drive carephilly View Post
Nice post.



A 4% increase in white people and <1% drop in the poverty rate

ohmygod!!!! displacement!!!!

Um, I have a pseudo-hypothesis that there was substantial displacement of (unsubsidized) childless adults, while (subsidized) poor families faced little displacement.

Since a childless adult at $12K income is defined as not poor, substantial displacement of childless adults typically would have little effect on the poverty rate.

I know a number of childless adults in Portland who have been displaced; since all of them are white, a whole lot of white people must have moved in to create a 4% increase in white people.
Reply With Quote Quick reply to this message
 
Old 02-12-2015, 09:23 AM
 
33,046 posts, read 22,053,448 times
Reputation: 8970
Quote:
Originally Posted by drive carephilly View Post
True. I've made this point a few times in different threads. That this sort of thing is relatively rare and mostly happening in the most expensive cities with a higher than normal % of renters. SF is a pretty extreme example because of the geography.

It's clear that things are changing in a lot of cities but, for the most part, the changes a lot more subtle than they're made out to be and in a lot of cases critics are pointing to rapid changes in demographics brought on by infill development as evidence of displacement. When you have 700 middle class white people show up in a new apartment complex built on a former parking lot in a neighborhood that used to be home to 500 people it changes the stats quite a bit.



I've actually done quite a bit of that going back to the 1940 census. Without interviewing people you can never know exactly why people moved but the data will make it obvious that certain things are or aren't happening.

I'm not really drawing conclusions from this single data point. I'm looking at the two together - 4% is kind of a joke in the context of the hair-on-fire displacement argument. Especially when you're looking at very modest declines in poverty. I recognize that the last decade has been a bad one for poverty but I'm also putting this in the context of all of the other data I've seen.

How Governing got it wrong: The problem with confusing gentrification and displacement | City Observatory




Boom.

I don't think 4% is kind of a joke. Figure 50% as a ballpark estimate of the proportion of homeowners in a neighborhood or census tract. It's the renters who are being displaced and replaced, and not so much the homeowners. So I think of that 4% as 4% of 50%, or 8% of the renters.

The (relatively) higher income renters are not being displaced, say one-third is a ballpark estimate of these renters, and two-thirds are lower income. So now we're up to 12% among lower income renters.

That is hardly a joke.
Reply With Quote Quick reply to this message
Please register to post and access all features of our very popular forum. It is free and quick. Over $68,000 in prizes has already been given out to active posters on our forum. Additional giveaways are planned.

Detailed information about all U.S. cities, counties, and zip codes on our site: City-data.com.


Reply
Please update this thread with any new information or opinions. This open thread is still read by thousands of people, so we encourage all additional points of view.

Quick Reply
Message:

Over $104,000 in prizes was already given out to active posters on our forum and additional giveaways are planned!

Go Back   City-Data Forum > General Forums > Urban Planning
Follow City-Data.com founder on our Forum or

All times are GMT -6.

2005-2019, Advameg, Inc. · Please obey Forum Rules · Terms of Use and Privacy Policy · Bug Bounty

City-Data.com - Archive 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35 - Top