[PRCo] Re: population trends

Mon Apr 9 15:14:42 EDT 2012

Depends on what we wish to use them for.   A lot of things we do have limited importance to anyone other than politicians.
1.   Consider national and state and municipal boundaries.   Do they have any importance except to politicians?   They are there so that the political dude has a ring with ropes around in which he can claim he is the main dude and no one else is allowed to step into his ring without his permission.  Frankly, John, I don't care of a Mexican steps into this ring looking for a job but the local political freak can use that guy as a way to get reelected.   He needs the stage with ropes around it in order to get reelected.  

Funny that most Indian tribes had a lesser need for absolute boundaries than our politicians do.  

Perhaps its all about money.  If your ring contains coal or oil or gas (a weak subject today), or a taxable factory, then we need to have those ropes in order to fight.   I frankly don't care.   

2.  The census?   Kindly read the U. S. Constitution, article 1, which prescribes that an enumeration shall be conducted within three years after the first meeting of the congress of the united states and every ten years thereafter as means of allocating seats in the house of representatives.   It was also to be used as a means of taxing the states. 

Federal and state policy has historically be based on the census.   Can you imagine the civil rights laws of the 1960s working without knowing how many minorities lived in different communities?   Remember that those laws required the hiring of blacks and other minorities in proportion to their numbers in the community.   Therefore we need the census to tell us how many Blacks, Spanish, Asians and so forth are living amongst us.  

Without researching every single law to determine how many required census data, remember that until and including the 1970 census, the long form (10% sample) asked if you had indoor plumbing or an outhouse.   That may have had a lot of importance to urban renewal projects in many cities.   I could see it even today being used to justify environmental legislation.    

However, since then we have found a lot of other reasons for the census and they are not entirely political.   For example, industries and sales firms and distribution companies require per-capita income date from the census in order to determine how to market their products.

Unfortunately, since so many people refused to fill out the census long form, the census department has replaced it with an on-going American Communities Survey.   I guess rather than hauling people into court who refused to complete a legally required survey, its better for the government to say, we don't know how to get good data?     The problem is, generally the better educated participate at higher ratios than the uneducated.   So if we don't force people, then the results of the survey will show that we have higher incomes, more education, more professional jobs than we really do have.   The whole thing becomes useless.

     http://www.census.gov/acs/www/about_the_survey/american_community_survey/

I am concerned if the national reliability is calculated at 0.1%, that there is virtually no validity at the local level.    The old census long form, which was given to 1 in 10 households and was required, gave really decent data about or cities and some rather respectable information even down to the township level.   I remember looking at the census data from 1970 for Camden County, New Jersey for commuting to work patterns.   Of course it was a 10% sample ... the long form ... and it was self identification so not everyone agreed on how they got to work ... but if you added together the number of people who claimed they went to work by train and those who went by subway and those who considered it a trolley car, you got nearly 40,000 people.   PATCO's riding counts were about 40,000.     I would not expect anywhere near that reliability today if the national numbers are assumed to be 0.1%.  
     http://factfinder2.census.gov/faces/tableservices/jsf/pages/productview.xhtml?pid=ACS_10_5YR_DP02&prodType=table

Here is a table from my township ... 38,000 people in the township ... in some categories the calculated error rate is 12 to 16 percent.

     http://factfinder2.census.gov/faces/tableservices/jsf/pages/productview.xhtml?pid=ACS_10_5YR_DP02&prodType=table

On Apr 9, 2012, at 9:00 AM, John Swindler wrote:

> 
> Perhaps much of the US census statistics are not important - except to politicians looking for someone to tax, and for playing games with number of congressional districts.   Don't overlook what Stasey told Geissenheimer back in 1972 - that about 25% of the voting rolls have changes each election.  People move.  They get jobs.  They get married.  They have children.  They buy homes.  Children get old and leave the home.  They retire.  They die.  So why should this all occur in the same county or even the same municipal district??       > Subject: [PRCo] Re: population trends 
>> From: fwschneider at comcast.net
>> Date: Sun, 8 Apr 2012 16:18:07 -0400
>> To: pittsburgh-railways at dementix.org
>> 
>> Some of the U. S. Census Bureau's intercensal estimates have been way off base.   Do I believe growth in Allegheny County?
>> 
>> The county is much more urban than the surrounding areas.   It includes Pittsburgh.  Many other cities showed unprecedented growth in the 2010 census ... perhaps we are learning that we cannot afford to live on huge lots in the suburbs in humongous homes that cost a fortune to heat and cool and require inordinate expense to get to and from our daily destinations.    Would I believe a slight gain ... maybe.   The gain they are showing is 0.2 percent, which, if extrapolated over ten years, would only be one-third of the loss between 2000 and 2010.
>> 
>> For the first time, Pittsburgh's unemployment is slightly below the national average.  Maybe the region has finally reached equilibrium and can accept a slight influx of people.  Time will tell.   
>> 
>> But the story tells us it was driven not by more births than deaths but by people moving into the area.   If you start with a given ... given we already believe that the population is growing, and we know from vital statistics that deaths exceed births, then we must blame the increase on people into the area.   Yes, you all know I am a cynic.   But I also know there is no good way to document migration between states or counties or cities.   
>> 
>> We should be creating the estimate by adding together births minus deaths plus in migration minus out migration.  However, I want you to tell me how you are going to measure migration from state to state.   If the federal government was really good at it, then we would know where all those Mexicans are dispersed!   :<)   Truth is, they don't know.   If they want to take the time to look, for example, at where courtesy claims for unemployment insurance are being filed, they might have some idea that people from Pittsburgh moved to Topeka or that people from Wichita moved to Dallas.   They might also get a clue by looking at school enrollment data.   But my experience in looking at some of their intercensal estimates makes me believe they are more along the line of projections based on the past than honest attempts at estimating the future.   The latter takes too much work and is awfully hard to defend.    But this Pittsburgh number is the reverse of the past.   I have no cl!
> ue!
>>  what they are doing.   Maybe they know some cities went up and think it's only proper to move them all up?????
>> 
>> I remember a urinating contest I got into back in the early 1970s over how many Spanish speaking people lived in Lancaster County. I inflamed the Spanish speaking community by telling them that my estimate was 2,500.  My estimate was one-quarter of the number they wanted us to believe.   I had based it on the percentage of kids in the schools and the family size of Spanish kids compared to non Spanish.   All knowns.   When the census came in a few months later at 2,475, we were of course both idiots ... I didn't know how to estimate and the census didn't know how to count.    But the guy who argued most loudly with me moved back to Puerto Rico.  
>> 
>> 
>> 
>> 
>> On Apr 8, 2012, at 10:36 AM, Dennis F Cramer wrote:
>> 
>>> http://www.pittsburghlive.com/x/pittsburghtrib/news/pittsburgh/s_790377.html
>>> This was in the Sunday (4-8) Tribune Review. What is not included in the 
>>> online version is the graph showing the various counties. I have attached a 
>>> scan of it.
>>> 
>>> Here is a small portion of the article.
>>> "The 10-county area of Western Pennsylvania showed population gains in 2011, 
>>> according to Census Bureau population estimates released last week.
>>> 
>>> Allegheny County's population increased by 2,233 people from 2010 for a 2011 
>>> population of 1.2 million people. The region's population - despite losses 
>>> in some counties, like Westmoreland, which saw a 614 decrease from 2010 - 
>>> rose by 930 people for a 2011 population of 2.6 million people.
>>> 
>>> The 10-county region includes Allegheny, Armstrong, Beaver, Butler, Fayette, 
>>> Greene, Indiana, Lawrence, Washington and Westmoreland counties.
>>> 
>>> Newcomers rather than newborns made the difference. The region had 3,468 
>>> more deaths than births, Census figures showed."
>>> 
>>> 
>>> 
>>>         Dennis F. Cramer
>>> http://home.windstream.net/dfc1
>>> 
>>> 
>>> 
>>> -- Attached file removed by Ecartis and put at URL below --
>>> -- Type: image/jpeg
>>> -- Size: 433k (444399 bytes)
>>> -- URL : http://lists.dementix.org/files/pittsburgh-railways/population%20trends.jpg
>>> 
>>> 
>>> 
>> 
>> 
>> 
> 		 	   		  
> 