Hacker News Reader

Comment by asift | Hacker News Reader

Your claim is sweeping because you're taking the results from one particular case study and extrapolating it out into a generic claim. It's overly simplistic because you're not acknowedging the reality that these conditions could change or may not be applicable in other situations (assuming there were no flaws in the underlying case study analysis to begin with).

Edit: this comment was made in response to toomuchtodo's comment prior to editing stating their claim was neither sweeping nor overly simplistic

View on HN

But isn't that how evidence in support of these these programs should be generated?

If not, how else can these programs be accurately evaluated prior to implementing on a more generic scale?

The idea is to find better solutions than we have now. The counterarguments you are raising sound to me like they could equally be applicable to any proposed solution. If we can't seriously consider something until we know it works in exactly the same conditions, I feel like nothing would ever change.

Let me know if I've read you wrong though.

There's absolutely nothing wrong with pointing to empirical evidence to support a claim (in fact, that is an excellent thing to do!). The issue I have with the original claim is that it isn't proven that giving away housing is a cheaper long-term solution than other forms of welfare. I certainly don't think we need to wait until something is proven to have an opinion or move forward in trying to make the world a better place (if we did, we would never take action in most cases), but we shouldn't present research in a way that disingenuously overstates our knowledge.

Cherry picking is allowed in contentious political questions.

Apparently.

This is usually a fair point to make, but you're pointing at neither the flaws in the underlying case study nor the weakness in generalizing Utah to a wider area. That's not really fair, especially since parent comment is providing data.

The parent comment is now significantly different from when I replied to it.

The articles were originally cited, but my response was in reference to their claim that their prior statement:

>its been proven that its cheaper to give the homeless housing than to pay for the resulting issues of not providing housing

...was neither sweeping nor overly-simplistic. The merits of the referenced case study--which may be an excellent case study--provide no grounds for refuting my claim that their prior statement was sweeping and overly-simplistic. Also, when making an argument based on extrapolated data, the burden for justifying extrapolation lies on the person making that argument.

Regardless, my only real point here is that the original claim (that giving people housing is cheaper than other forms of welfare has been "proven") is simply too powerful of a claim. It might be a good policy and it may in fact be cheaper, but it's incorrect to say any position on this topic has been "proven".

asift,

I did indeed edit my comment, removing my assertion that my claim was not sweeping and overly-simplistic (I wish HN would keep an edit history that was visible).

With that said, I still disagree with you. If an experiment takes place, and data is apparent, that data stands on its own merit; if that data can then be used to make a statement, its proof, not "sweeping and overly-simplistic".

You can disagree with me, that's fine. I believe, based on the data, that we should be giving housing to the homeless. If you don't agree, I encourage you to engage politically; I do, and I'm passionate about positive social change. I appreciate you engaging in discourse with me in a civil manner; it happens much less than I wish it would (in general).

I'm fairly new to HN, but I've found people here tend to be much more civil than most places on the web, which I think is great.

>if that data can then be used to make a statement, its proof

I'm not trying to be overly semantic, but I think "evidence" would be more appropriate than "proof" in this case. My claim was never that there isn't evidence that housing the homeless costs less than other welfare programs (or that I think it's something we shouldn't do), I just think you are being too forceful in your claim.

A community in Alberta, Canada has, in effect, ended homelessness [0]. The mayor claims[1] the cost of providing housing to someone is around CAD20,000, but the cost of a homeless person can be up to CAD100,000 - meaning that it's financially prudent for the municipality to proactively house people.

Both links are to press reports and don't provide any further information to back up the claims, but given that Medicine Hat and several other municipalities have taken up objectives of ending homelessness, I suspect it wasn't just for the warm fuzzies of giving people roofs over their heads. A lot of people (including the Mayor himself!) would have the same knee-jerk "they didn't earn it" response.

[0]: http://www.cbc.ca/news/canada/calgary/medicine-hat-on-brink-...

[1]: http://www.cbc.ca/radio/asithappens/as-it-happens-thursday-e...

Thank you for contributing an additional data point!

There is a big moral hazard issue here as well. When you are working for minimum wage to rent the cheapest apartment in town why not become 'homeless' and get a house for free. I have no doubt that if you could eliminate the frauds from the system it would be cheaper, but I don't think you could.

What is a "fraud"?

The simple way to solve this is to offer housing free to all comers, but make it less attractive than existing options.

I'm sure logic like that works well on the playground and at the grateful dead concert, what about all the people who invested in property? What about the value of homes in general? Where would you even house all these people? It isn't always so simple.

If homes lose value, so be it. And where would we house all of these people? ~500K people in the US are homeless [1], providing housing for 1/600th of the US population is not difficult.

[1] http://endhomelessness.org/library/entry/the-state-of-homele...

Well, do you buy property to live on it, or to "invest"?

Real estate speculators hardly seem to add any value to the economy.

Why do we care about people who treat property as an investment, rather than actually using the property?

Reconsider preconceptions of what cities and society can do for humanity.

Basic needs like clean water, sanitation, and streets could be extended to public housing and food. Public housing might be a bare minimum room in a crowded building on cheap land around urban areas with public transit connecting the community into the city.

1: Not everyone lives in cities. 2: Why am I paying for your housing and then having to either live in a slum, or buy one myself?

1: Correct, though since the majority of jobs, infrastructure, and population gravitates around urban areas, building further infrastructure near populated areas would be reasonable.

2: Why do societies agree to build and maintain public infrastructure projects like sanitation, water, roads, schools, etc?

There will always be the economy of WANTs, buy a house wherever you want!

Consider the excessive amount of tax money going into the Prison Industrial Complex economy, it seems a much better option to build low cost / public housing infrastructure and job training rather than prisons.

1: So, basically people who live in rural areas will subsidize the people who live in cities? Or will they just pay much lower taxes?

2: Sanitation? I have my own septic system.

Water? Well.

Roads? You mean the gravel thing that has tons of pot holes? It would not be feasible for me to do that alone. (private property and all)

Schools? Tell me how much a high school degree gets you in today's economy? I just don't see much ROI there.

I agree about the prison system, it should only have violent criminals and thieves but the savings should go to lower taxes for the poor and middle class.

I don't see any moral hazard. If anything, not providing housing for those who don't have it is the morally hazardous path.

Reply on news.ycombinator.com