Census Bureau To Scrap Handhelds — Cost $3 Billion

GovTechGuy writes "The Census Bureau will tell a House panel today that it will drop plans to use handheld computers to help count Americans for the 2010 census, increasing the cost for the decennial census by as much as $3 billion, according to testimony the Commerce Department secretary plans to give this afternoon."

Comments

Surplus

[Jhon]

Will they sell the hand-helds? Or give them away like Cheese in the 80's?

Re:Surplus

[SQLGuru]

Look for them on Woot during the next Woot-off.

Layne

Re:Surplus

[d3ac0n]

I don't think they ever bought them. The cost difference is related to the extra time and manpower that a paper census will take vs the costs for an electronic one.

Personally I think this is a good thing. Better to spend money to do things the tried and true way than to experiment with a "hi-tech" solution that may or may not have exploitable weaknesses in it. We've all seen how faulty the electronic voting machines have been, I think it's wise that the census folks don't want to go down that road.

Kudos to the Census people, and to House Speaker Dennis Hastert (R-Neb) for supporting and encouraging their wise decision.

Re:Surplus

[Gat0r30y]

I don't think they ever bought them. Nah, didn't buy em, just dropped 1/2 a Billion into development it would appear.
From the Article -

In 2006, the Census Bureau awarded a $595 million contract to Harris Corp. to develop more than 525,000 handheld computers that enumerators would use to collect data from Americans who did not send in their census forms.

Re:Surplus

[MightyMartian]

Generally when you are trying out new technology, you choose a few locales to be testbeds. That way you can determine whether or not the technology will work as advertised, and if it does, it gives you a chance to correct any bugs. To go out and buy three billion dollars worth the equipment and then decide that it doesn't work suggests to me that there are some severely incompetent people at the top of the chain.

I feel the same way about voting machines. Test them out in a few places, get to know the equipment, and if you still figure it's going to work, you have a place to go. But this mass exodus from one system to another is just lunacy.

Re:Surplus

[$RANDOMLUSER]

I'm sure it comes as a surprise to former House Speaker Dennis Hastert that he's from Nebraska - since he's from Illinois.

Re:Surplus

[timeOday]

Personally I think this is a good thing. Better to spend money to do things the tried and true way than to experiment with a "hi-tech" solution that may or may not have exploitable weaknesses in it.

I can't imagine WalMart, or any other successful business attempting to do inventory (yes, that's what a census amounts to) purely on paper because they can't get their act together, or have money to burn. This is just as frustrating as the IRS refusal to offer an official free tax filing website. $3 billion extra dollars! All for a census that's riddled with extra transcription errors and will obviously be entered into computers in the end anyways, to be of any use at all.

Promise and risk of electronic census.

[gnutoo]

I've done a census and think GPS enabled devices would greatly increase accuracy but it will also greatly increase costs. A sad fact is that people don't really go all the places they are supposed to go and honest enumerators don't last long in places that stick to quotas. GPS and time tracking devices will prove that the enumerator actually visted each and every place they should have. A mashup with something like Google maps will show if areas have been neglected. An honest census will take significantly more manpower than the one we have now.

There are, of course, the same kinds of risks we have seen with electronic voting. The only solution is to be as transparent as possible. Non free software is a no-no.

It's all borrowed anyway...

[bstarrfield]

Just another example of the mind boggling inefficiency and ineffectiveness of the current American administration. $3 billion dollars would cover roughly a week of expenses in Iraq - so the sum must be inconsequential.

Or - $3 billion dollars could pay for the college tuition of thousands of students, could dramatically raise NSF funding, or could help rebuild our roads. Don't these people even shame anymore?

One of the fun points about this is that the current Administration was elected (partially) on their supposed business expertise. Which appears to be actually true as many major businesses flub their own large scale IT projects.

Well - given that we're running a fantastic deficit, we'll just throw the extra costs of the the census project into our staggering debt.

$10/person ?!?

[vijayiyer]

It costs $10 _per person_ to count us? That's unbelievable. Perhaps if they just count people (as the Constitution requires) rather than gather race and demographic information, they could cut their costs.

Re:$10/person ?!?

[trooper9]

It costs $10 _per person_ to count us? That's unbelievable. Perhaps if they just count people (as the Constitution requires) rather than gather race and demographic information, they could cut their costs. If they did that, there wouldn't be enough information to allow groups to claim "victim" status for whatever social variable they perceive that sets them apart. Remember, the census does more than count, it helps us cordon-off certain groups on our Level Playing Field.

Re:$10/person ?!?

[eln]

I bet if they just gave everyone $5 as an incentive to self report, you could get more accurate results at half the cost.

Re:$10/person ?!?

[Unordained]

No. The cost is increasing by $3 billion (with a b). From the article:

Gutierrez said reverting to a paper-based census, in addition to other costs not associated with the handhelds, is expected to increase the cost of the 2010 census to between $2.2 billion and $3 billion through fiscal year 2013. That would bring the total cost of the 2010 census to between $13.7 billion and $14.5 billion. He said the bureau would need an increase of $160 million to $230 million for fiscal 2008 to cover costs associated with returning to paper, with an additional $600 million to $700 million for fiscal 2009. Gutierrez added that the majority of the cost increases would occur in 2010

So it actually costs somewhere around $37/person to count and classify each of us, or around 7 hours of minimum-wage labor. It's far worse than you think.

Also, the handhelds were for field operatives collecting data from people who didn't send in their forms -- the cost estimate above includes the distribution and processing of paper forms that you fill out yourself, which you could reasonably expect to be cheaper than going door-to-door collecting data, thus increasing the per-person cost of personal data collection.

OMG! That's 3 days of Iraq war spending!

[Reality Master 201]

Stop the waste now!

Re:Bzzzt, wrong!

[d3ac0n]

Actually, it is highly likely that it is a bit of both. From the article:

At a March 5 hearing of the Senate Homeland Security and Governmental Affairs Committee, Gutierrez said, "significant miscommunication concerning technical requirements between the Census Bureau and Harris" were a main reason for the failings.

I think it was a situation of the "Left hand not knowing what the right hand is doing". Not an uncommon problem in both corporate AND governmental circles. Having previously worked for a company that dealt with government contracts, I can say without a doubt that it is pretty much par for the course when doing that type of work.

I'm just glad to see that the Independant panel had the good smarts to decide to just scrap it and go back to the old way. I can't imagine how much money would have been wasted trying to implement things as they were. Good-on them.

Re:Bzzzt, wrong!

[azadrozny]

He points to a dress rehearsal held in May 2007 as when "development and scoping problems emerged." The bureau then identified "more than 400 new or clarified technical requirements," he said, which were delivered to Harris on Jan. 16.
It appears that the government shares some of the blame. 400 new/modified requirements tells me they didn't have good idea of what they needed the system to do. A system is only as good as the specification provided.

Re:Census? Just count me out.

[stranger_to_himself]

The thing is, race does matter, and you can't make racism go away by pretending it doesn't exist, or saying it shouldn't exist (which of course it shouldn't). Issues do affect different racial groups in different ways. By denying this you prevent the application of solutions where problems arise, making them far worse.

Re:Census? Just count me out.

[Immortal Poet]

If you're not racist, then race should NOT matter at all. Anything less makes you racist to some degree. It is absolutely ludicrous to purport that someone becomes a racist by acknowledging the mere existence of race. Being aware of the differences between different groups of people is one thing; believing those differences makes one group better than another is what racism really is. Don't be naive enough to get the two confused.

Re:Census? Just count me out.

[trolltalk.com]

Why is that so upsetting to you? Gross estimations of your ethnicity can be made by simply looking at you; if you leave your house, your ethnicity is essentially public information, right?

Let's see ...

When you buy something at the store, you're standing in line with other members of the public, so your purchases are essentially public information, right?

When you take a book out of the library, your reading tastes are essentially public information, right?

When you visit a hospital or clinic and are sitting with strangers in a waiting room, your medical problems are essentially public information, right?

When you take a book out of the library, your reading tastes are essentially public information, right?

When you pick a dvd off the shelf to rent, your viewing interests are essentially public information, right?

When you shop for groceries, your eating habits are essentially public information, right?

When you buy a present for that someone special to surprise them, your purchase is essentially public information, right?

So, where do you draw the line?

For example, finding out that one county's minority population is 13% below the poverty line, while another county's rate is only 5%. It would be useful to know that situation even exists; then, you can try to find out what the difference is and try to help the situation.

There is a happy medium between affirmative-action-type policies and nothing. It is useful for sociologists to have this kind of information.

So you would make it that aid to help people escape poverty should be targeted by skin colour, rather than need? Come on, poor is poor - when you're broke, hungry, and homeless, your skin colour doesn't make your stomach growl any less.