Census Projections - Rocky Mountain States

[muon2]

I've added the Rocky Mountain states to the links from the 2010 apportionment page. As before, comments or questions are welcome.

[jimrtex]

I've added the Rocky Mountain states to the links from the 2010 apportionment page. As before, comments or questions are welcome.

It is surprising that there is only one split needed in Colorado with 4 large counties together.

BTW, there are parts of Arapahoe County that are surrounded by Denver, so this ends up being a split as well.  But this would only require an adjustment in the boundary between 1 and 7.

[Padfoot]

beautifully done as usual muon2.  It will be interesting to see how close you are to accurately projecting the congressional maps, especially Nevada.

[muon2]

I've added the Rocky Mountain states to the links from the 2010 apportionment page. As before, comments or questions are welcome.

It is surprising that there is only one split needed in Colorado with 4 large counties together.

o
If I want to divide 4 roughly equal counties into 3 districts, I have two choices to minimize splits. I can split one county three ways, adding a piece to each of the three intact counties. Or I can split two counties two ways each, adding one piece from the two split counties to each of the intact counties, then attaching the other split pieces together. Since I view this latter style as creating four partial counties in districts, I prefer the former since only three partial counties are created. As I note in my description, I do provide some deference to keeping incorporated places intact, and that can modify how I prefer to split counties.

BTW, there are parts of Arapahoe County that are surrounded by Denver, so this ends up being a split as well.  But this would only require an adjustment in the boundary between 1 and 7.

Glendale and a couple of unincorporated patches amount to a few thousand people surrounded by Denver. I can estimate Glendale, but the Census doesn't provide anything to estimate the other areas isolated in Arapahoe.

There are a few places in the US where part of a county is detached from the rest (eg. Brookline MA). I've never really considered those to be a split, since contiguity is a foremost requirement of a district and to attach pieces often requires splitting another county instead, as it would in CO.

[muon2]

beautifully done as usual muon2.  It will be interesting to see how close you are to accurately projecting the congressional maps, especially Nevada.

The real Nevada map will almost certainly have a three all or primarily Clark Co districts and a Reno-Tahoe-Carson City district. Which low population desert areas attach to the Clark district(s) may well vary from my map. I don't know if NV has a rule prefering intact counties, and slight shifts in the actual census numbers will change which small counties best make the population match.

Also, the Reno district could swap Elko and Humboldt in the northeast for Mineral, Esmeralda, and Nye in the southwest. This would allo the adition of a little bit of Clark county and get the population counts of all districts to the exact ideal value. Clark gets split into two whole and two partial districts, and there would be some splitting of municipalites to get the exact count.

[jimrtex]

BTW, there are parts of Arapahoe County that are surrounded by Denver, so this ends up being a split as well.  But this would only require an adjustment in the boundary between 1 and 7.

Glendale and a couple of unincorporated patches amount to a few thousand people surrounded by Denver. I can estimate Glendale, but the Census doesn't provide anything to estimate the other areas isolated in Arapahoe.

4547 in Glendale
2517 in Holly Hills
41 in other areas.

7105 total.

Holly Hills is mostly single family residential, and Glendale is mostly businesses with apartments.  If there are any estimated increases for Glendale, it would be based on construction of new high rise apartments, it would not be something to project forward to 2010.  The census estimate for Glendale in 2005 was 4771.  So I would use 7105 for the projected population, and assume all increases occur elsewhere in Arapahoe county.

[Cubby]

Great work on both these series Muon!

Don't you think having Douglas County and Boulder County in the same district could lead to a culture clash? The former has more in common with El Paso County, but by 2010 they will be too large to form 1 district. Boulder should go with Jefferson or Larimer and some smaller counties.

Do you think Idaho will have a third seat in 2020? How would that look, assuming the growth continues in the same places? 

[jimrtex]

Don't you think having Douglas County and Boulder County in the same district could lead to a culture clash? The former has more in common with El Paso County, but by 2010 they will be too large to form 1 district. Boulder should go with Jefferson or Larimer and some smaller counties.

Longmont, Lousville, and Lafayette are not that dissimilar to Douglas County.

It is a feature of a system that avoids splitting counties that political considerations are ignored when drawing the districts.

Do you think Idaho will have a third seat in 2020? How would that look, assuming the growth continues in the same places? 

It will be close.  In 2000, Idaho was almost exactly at 2/435 of the US population.  To get a 3rd seat, it would need around sqrt(6)/435 of about 2.45 / 435 of the population.  That would require a growth of 22% faster than the US growth rate or around 40%.   Adding 500,000 to 1.3 million in 20 years is a healthy increase unless you're Nevada.

[muon2]

Don't you think having Douglas County and Boulder County in the same district could lead to a culture clash? The former has more in common with El Paso County, but by 2010 they will be too large to form 1 district. Boulder should go with Jefferson or Larimer and some smaller counties.

Longmont, Lousville, and Lafayette are not that dissimilar to Douglas County.

It is a feature of a system that avoids splitting counties that political considerations are ignored when drawing the districts.

Exactly as jimrtex says.

Do you think Idaho will have a third seat in 2020? How would that look, assuming the growth continues in the same places? 

It will be close.  In 2000, Idaho was almost exactly at 2/435 of the US population.  To get a 3rd seat, it would need around sqrt(6)/435 of about 2.45 / 435 of the population.  That would require a growth of 22% faster than the US growth rate or around 40%.   Adding 500,000 to 1.3 million in 20 years is a healthy increase unless you're Nevada.
[/quote]
The Census projections to 2020 were relased in April 2005. Based on those ID only would have a population equal to 2.26 of a seat in 2020, so they would stay at 2 seats. However, if the current growth rate is maintained, the ID population will be much closer to 2.5 seats in 2020, and the state would get a third seat.

[MaC]

I'm surprized North Carolina isn't gaining any seats-wasn't that previously a fast growing state?

[Cubby]

I'm surprized North Carolina isn't gaining any seats-wasn't that previously a fast growing state?

It gained a seat in 2000, so it will have to add another 1,500,000 people to get one more seat. Based on 2006 estimates, it will add about 1,290,000.

But as we saw in 2000, the Census Bureau is comically inaccurate, so there's a decent chance North Carolina could get another one.

[jimrtex]

I'm surprized North Carolina isn't gaining any seats-wasn't that previously a fast growing state?

It barely got the 12th seat, and that was with favorable rounding.  This decade it is consolidating its gains, and will get a 13th in 2020.

[Tender Branson]

I´m pretty sure Idaho will get a 3rd seat after 2020. Growth picked up speed in the period after 2005 and the state is now the 3rd fastest growing state in the nation. I think ID will grow an average 2.5% in the 2000-2010 span (defying the low Census Bureau estimates of 2.0% growth) to reach 1.66 Mio. people and growing further to close to 2 Million in 2020. A third seat is almost guaranteed than when Nebraska is overtaken by Idaho in its population. Nebraska now has 3 districts.

[muon2]

I´m pretty sure Idaho will get a 3rd seat after 2020. Growth picked up speed in the period after 2005 and the state is now the 3rd fastest growing state in the nation. I think ID will grow an average 2.5% in the 2000-2010 span (defying the low Census Bureau estimates of 2.0% growth) to reach 1.66 Mio. people and growing further to close to 2 Million in 2020. A third seat is almost guaranteed than when Nebraska is overtaken by Idaho in its population. Nebraska now has 3 districts.

That's a pretty big difference going from 2.0 to 2.5%  this decade. When you claim that increase are you expecting that the growth has already been higher in the first half of the decade, or is the second half going to be at or above 3.0% bring the net to 2.5%? In either case do you have any source material from the state or county to suggest the difference from the Census?

[Tender Branson]

I´m pretty sure Idaho will get a 3rd seat after 2020. Growth picked up speed in the period after 2005 and the state is now the 3rd fastest growing state in the nation. I think ID will grow an average 2.5% in the 2000-2010 span (defying the low Census Bureau estimates of 2.0% growth) to reach 1.66 Mio. people and growing further to close to 2 Million in 2020. A third seat is almost guaranteed than when Nebraska is overtaken by Idaho in its population. Nebraska now has 3 districts.

That's a pretty big difference going from 2.0 to 2.5%  this decade. When you claim that increase are you expecting that the growth has already been higher in the first half of the decade, or is the second half going to be at or above 3.0% bring the net to 2.5%? In either case do you have any source material from the state or county to suggest the difference from the Census?

What I´m assuming here is that the Census Bureau Population Estimates are again lower than the actual Census Count in 2010 will be.

For example the April 1, 2000 official Census count showed the population to be higher by 2.5% than the Census Bureau Estimates for that date.

I´m now assuming:

Idaho`s population was 1.294 Mio on Census Day April 1, 2000.

The Census Bureau Estimate for Idaho on July 1, 2006 was 1.467 Mio., indicating a 2.02% growth rate in the last 6.25 years.

But in 2005-06, growth in Idaho picked up speed to 2.6%, which I assume will continue to April 1, 2010 bringing the population to 1.615 Mio.

Growth then was 2.24% for the 2000-2010 period.

1.615*1.025 (the continued overcount from the 1990-2000 period I assume) = 1.655 Mio. = 2.5% annually from 2000-2010.

[muon2]

I´m pretty sure Idaho will get a 3rd seat after 2020. Growth picked up speed in the period after 2005 and the state is now the 3rd fastest growing state in the nation. I think ID will grow an average 2.5% in the 2000-2010 span (defying the low Census Bureau estimates of 2.0% growth) to reach 1.66 Mio. people and growing further to close to 2 Million in 2020. A third seat is almost guaranteed than when Nebraska is overtaken by Idaho in its population. Nebraska now has 3 districts.

That's a pretty big difference going from 2.0 to 2.5%  this decade. When you claim that increase are you expecting that the growth has already been higher in the first half of the decade, or is the second half going to be at or above 3.0% bring the net to 2.5%? In either case do you have any source material from the state or county to suggest the difference from the Census?

What I´m assuming here is that the Census Bureau Population Estimates are again lower than the actual Census Count in 2010 will be.

For example the April 1, 2000 official Census count showed the population to be higher by 2.5% than the Census Bureau Estimates for that date.

I´m now assuming:

Idaho`s population was 1.294 Mio on Census Day April 1, 2000.

The Census Bureau Estimate for Idaho on July 1, 2006 was 1.467 Mio., indicating a 2.02% growth rate in the last 6.25 years.

But in 2005-06, growth in Idaho picked up speed to 2.6%, which I assume will continue to April 1, 2010 bringing the population to 1.615 Mio.

Growth then was 2.24% for the 2000-2010 period.

1.615*1.025 (the continued overcount from the 1990-2000 period I assume) = 1.655 Mio. = 2.5% annually from 2000-2010.

I follow everything except why you project a jump to 2.6% in 2005-2006. If the reason is to get 2.24% then a jump to 2.6% from 2.0% is very artificial. Also an underestimate in one decade does not imply a repeat. If anything, the Census will focus on techniques to compensate for the past error in the estimates this decade. If there is new unaccounted for growth from birth or immigration then your case would be stronger.

[Tender Branson]

I follow everything except why you project a jump to 2.6% in 2005-2006. If the reason is to get 2.24% then a jump to 2.6% from 2.0% is very artificial.

That´s because the Census Bureau estimate for 2005-06 showed Idaho picking up speed in its growth to 2.6%, which I expect to continue until April 1, 2010.

Link:

http://www.census.gov/Press-Release/www/releases/archives/population/007910.html

"Meanwhile, Arizona was the nation’s fastest-growing state over the period, breaking Nevada’s grip on the title, with its population rising 3.6 percent. Nevada ranked second this time,as its population climbed by 3.5 percent, followed by Idaho (2.6 percent), Georgia (2.6 percent) and Texas (2.5 percent)."

Also an underestimate in one decade does not imply a repeat. If anything, the Census will focus on techniques to compensate for the past error in the estimates this decade. If there is new unaccounted for growth from birth or immigration then your case would be stronger.

Immigration (most of it illegal) is hard to predict, but as I´ve said I just assume it will again be underestimates, which can be true or not. Census 2010 will show.

[jimrtex]

What I´m assuming here is that the Census Bureau Population Estimates are again lower than the actual Census Count in 2010 will be.

But aren't also assuming that the error will be in Idaho's favor?  That they are underestimating Idaho, more than some other state?

[Tender Branson]

What I´m assuming here is that the Census Bureau Population Estimates are again lower than the actual Census Count in 2010 will be.

But aren't also assuming that the error will be in Idaho's favor?  That they are underestimating Idaho, more than some other state?

I´ve also checked that, according to the states population estimates from 1990-2000. There was no significant difference in the underestimation in the US and Idaho. Both in the US and Idaho the Census count showed the actual population higher by 2-2.5% of the estimated population.

[muon2]

I follow everything except why you project a jump to 2.6% in 2005-2006. If the reason is to get 2.24% then a jump to 2.6% from 2.0% is very artificial.

That´s because the Census Bureau estimate for 2005-06 showed Idaho picking up speed in its growth to 2.6%, which I expect to continue until April 1, 2010.

Link:

http://www.census.gov/Press-Release/www/releases/archives/population/007910.html

"Meanwhile, Arizona was the nation’s fastest-growing state over the period, breaking Nevada’s grip on the title, with its population rising 3.6 percent. Nevada ranked second this time,as its population climbed by 3.5 percent, followed by Idaho (2.6 percent), Georgia (2.6 percent) and Texas (2.5 percent)."

Your hypothesis is used by some projections. They usually take the average of the last three years and use that rate to project forward. ID has had estimates of 2.5% and 2.6% in the last two years after growth rates from 1.66% to 1.98% at the beginning of the decade.

My method assumes that individual year fluctuations average out, while you assume that the latest year reflects the real trend. A case can be made for either method, but since the estimates quoted annually are just estimates, I tend to be cautious in giving any one year too much weight.