Nov 7, 2016 "fact": Hillary Clinton has a 98.2% chance of winning the presidency: http://www.huffingtonpost.com/entry/polls-hillary-clinton-wi...
There are facts, but it does them no service to overstate what the evidence actually says.
There are facts, but it does them no service to overstate what the evidence actually says.
But the types of statistics here are fundamentally different. If you look at the huffington post numbers as how many people will vote for a candidate, then even their prediction is pretty close to correct. A few million votes make the difference between 50% and 99%, but a few million votes is a reasonably tight accuracy range for "how many". Similarly I would expect something like the 80% number to be pretty close to correct, at least based on the definitions chosen.
The probability of dying in a car crash is way less than 1% and still some people are unlucky enough to die in this way every day. It's a matter of sample size. In the same way, the vast majority of the time a candidate with a very slim chance won't make it to presidency, but it happened that Trump did.
I think people discussing complex phenomena (like poverty) tend to overstate the power of their evidence; they overstate its ability to support the conclusions they want to draw from it.
This was illustrated in dramatic fashion when many/most of the most educated, "enlightened" people in the US had a reality check that their model was off last November. If there hadn't been an actual vote, people would probably still be insisting that Hillary Clinton would have won, had there been an actual vote on Nov 8.
And that was the simple question of "how many people will show up to a polling station and mark a box for HC vs DT"! And the evidence consisted of pollsters who were asking hundreds or thousands of people that precise question every day! That should have been much easier to model and predict than poverty! It didn't even get into the causes of why people think or vote in certain ways.
And yet still we have people looking at complex systems and problems like poverty and saying "it's not because of X, it's because of Y" with a feeling of certainty. And thinking that they are justified in their certainty because they claim this is what the "facts" say.
I think, first of all, that grandparent hasn't even accurately summarized their own link, which says "Children and eldery make up slightly more than half of the market poor." Their statement that "essentially" 80% of the poor are children and elderly has already caused a significant amount of the underlying data to be "lost in translation".
I think, second of all, that bucketing elderly, children, the disabled, etc into "vulnerable populations" and saying they comprise 80% of the poor doesn't really support the author's thesis at all.
No one blames a child for being poor. It is not a child's responsibility to support themselves. So saying children are poor doesn't really speak to whether poverty is a personal choice. Children are the responsibility of their parents, so it is the parents choices that would be relevant here.
Elderly people, on the other hand, generally spend their lives realizing they will be old someday, and therefore poverty of an elderly person does not necessarily say anything about whether poverty is a personal choice. It could be a lack of effort put into saving during their working years.
I want to be clear that I'm not against safety nets, or in favor of doling out blame and judgement towards poor people.
What I am against is the (to me) crazy amount of overconfidence many people (on the left especially) have in the ability of their data to "prove" their political positions and arguments.
And I'm not even against people's right to hold convictions that the evidence doesn't fully justify. What I am especially against is when people use those convictions to pass blame and judgment on others. The clear subtext of this article is that we are all guilty, as a society, for making poor people poor.
Part of this statement seems to reflect a high degree of confidence in the idea that this is phenomenon "especially" afflicts the left, but there doesn't seem to have been a presentation of evidence which provides unambiguous support for the thesis.
> but there doesn't seem to have been a presentation of evidence which provides unambiguous support for the thesis.
I said "(to me)" as a disclaimer that I was offering my opinion/belief, and not a provable thesis.
Remember, n=1 in this scenario.
1. the specific circumstances of the outcome. Hillary Clinton didn't lose just barely, she lost by a lot. And she lost states like Wisconsin, which wasn't even supposed to be in play.
2. the fact that predictions like HuffPo's were far off from more credible pundits like 538.
3. the fact that HuffPo so aggressively mocked anyone who took a more measured view. http://www.huffingtonpost.com/entry/whats-wrong-with-538_us_...
The only difference here is that there's no "poverty election day" where the entire population goes to register their poverty level. So we never actually get to see if the prediction was correct!
Combine that with the data from the correspondinc Apr 15th, and we should have something pretty close to a "poverty election day".
[0]: cell phone and spotty network made it too hard to look up a good source - please feel free to add them if you can)