Charts and graphs are great, because they can let you see a pattern that you might not see in a spreadsheet, but they only work when you use the actual data. Fox News isn’t doing themselves any favors by putting up this chart. It shows the recently announced drop in unemployment rate to 8.6 percent as a non-change.
The November rate is lower than the March rate of 8.8 percent, but it’s shown to be higher in the Fox News chart. Here’s what the graph should look like, according to data from the Bureau of Labor Statistics:
It doesn’t quite beat the best pie chart ever, but it’s close.
Honest mistake? Wishful thinking? Little of both?
[Media Matters via @periscopic]
You know Fox News is meant to be unbiased, right?
It clearly is not but MSNBC stinks of liberal bias. I’m a lifelong Republican but I only read Reuters. We still lack a popular, balanced source of news in America. This chart certainly does not make Fox News reliable but at least it can be considered reliable considering its nature.
Nevertheless, there have been enormous increases in public debt–no doubt caused by Obama’s blue-collar public sector support. He’s causing more problems than providing solutions to be fair. Although he is one of the better Democratic presidents of modern times, there are better alternatives. I won’t give my preference and with this move I am allowed to put the message out there that FlowingData should not push a political agenda. I understand this is your personal blog on statistics, visualisation of them and more, but it should remain confined to this subject.
Ionnas, “no doubt”? There are a lot of doubts about that statement; I don’t think that “blue collar public sectorN has any real meaning. Debt started rising after Bush’s first two unpaid for tax cuts. The wars in Iraq and Afghanistan were also unpaid for. There was also the huge expansion of medicare which was also, you guessed it, unpaid for. Obama’s policies have added to debt, but keep in mind that times of high unemployment is the only time when deficits are justified. The main reason for the massive increase in debt is the lower tax revenue because of high unemployment and the Bush tax cuts. Take a look at this chart:
Your ability to judge bias is quite poor. You think FOX news is reliable because it’s trying not to be biased, but you think this post shows bias. You should reread this post and try to find where FlowingData injects any political agenda, other than saying that FOXnews often has bad charts, which we have empirical data to support (it’s the big picture at the beginning of the post) . You on the other hand, have not supplied data for your claims, “to be fair”, but have only provided opinions. Also, why do you get to tell someone what opinions they express on a blog?
Funny how debt only matters when a Democrat is president. When a Republican’s in the White House, as Cheney put it, “Reagan proved that deficits don’t matter.”
As for bias, the idea that anybody can think “MSNBC stinks of liberal bias” yet also claim “Fox News is meant to be unbiased”…that’s just nuts. Fox News is meant to be as biased as possible in favor of conservatives. That’s the channel’s sole purpose for existence, and always has been. The reverse is not and has never been true of MSNBC. NBC simply makes a half-hearted attempt at a primetime lineup of liberal commentators because CNN’s attempt to be Fox News Light has been such a ratings failure. But they always keep a leash on their liberals commentators to keep them from getting too liberal and thus hurting the parent company’s corporate agenda. Hence two of their most popular hosts getting run off for refusing to play softball.
You’ve got to be kidding in your assessment of MSNBC.
Also, do you intentionally take Cheney’s comments out of context or do you really have as little understanding of them as you seem to? The idea that “deficits don’t matter” does not exist in a vacuum. This concept refers to the debt/deficit to GDP ratio. The deficit has increased under every president of the last 40+ years. Under Reagan, Clinton, and Bush43 the GDP rose faster than the deficit so the debt to GDP ratio was less problematic. Under Bush41 this was not true. Though the debt did not increase as sharply, neither did the GDP. Under Obama, the deficit has increased dramatically and the GDP has gone down.
And don’t try to cite that chart about presidential debt that Pelosi put on her site. That is a bigger and more obviously intentional lie than this by leaps and bounds.
We have a pretty unbiased source in NPR. And what on earth do you mean that they are meant to be unbiased? That notion is purely Orwellian. Other outlets may reveal bias in less intentional ways but no outlet that claims to be mainstream is guilty like Fox of being almost purely a source of propaganda. No wonder we have polls that show Fox viewers knowing less than people who watch no news.
If you seriously believe that NPR is in any way, shape, or form and unbiased news source then you have much bigger problems than the chart makers at Fox.
Also, “no outlet that claims to be mainstream is guilty like Fox of being almost purely a source of propaganda”? Apparently you’ve never heard of msnbc? … or do they not fit into “claims to be mainstream”?
How about CBS? Check out http://www.ratherbiased.com for some examples.
Where’s MSNBC’s graph?
I dont see haw there is any bias… all they said was it was a graph that misrepresents the data.
You were the one that added the editorial about Fox News
There is no editorial against Fox News… other than saying they sometimes make bad charts and asking if it’s a mistake or “wishful thinking”.
I’ve never seen a bad chart defended on Flowing Data before, ever… Who are these bad-chart-loving commenters and where have they come from today?
A visualization that doesn’t show what the data represents is EXACTLY like a sentence that doesn’t represent the truth. Both of them are lies: one is a verbal lie, one is a visual lie. This is a non-partisan, apolitical truth. If it was by accident, then it could represent incompetence. If it was on purpose, then it could be bias. It doesn’t matter what organization put it out there, it’s either a lie or incompetence on the part of the visual designer.
Both of them are lies: one is a verbal lie, one is a visual lie. This is a non-partisan, apolitical truth.
This is extremely well said. I’d like to use part of this in my letter to Fox News. May I?
You may also notice that the 9.2 on the chart is a line-and-a-half above where 9.0 is, yet the 8.8 on the chart is only about a line below 9.0 — again making the numbers look inflated.
yeah, i think the 9.2 point is actually above the 9.5 mark on the axis. so crazy bad.
Interestingly, there are 5 data points listed on the Y axis but there are 9 horizontal lines across the body of the chart. That’s just bizarre.
Edward Tufte would not approve.
Can we please stop talking about Tufte as if he’s Jesus? Sorry, it’s nothing personal, but Tufte-thumping seems to be on the rise…
The problem with this chart isn’t that it doesn’t meet the exacting standards of purity needed for inclusion in Tufte’s hypothetical visualisation museum. It’s not trying to.
The problem is that it contains a blatant lie, and a lie which (coincidentally?) happens to match up with the publisher’s political line. “Don’t completely lie your arse off in a chart” isn’t the kind of difficult abstract principle that needs a Yale professor to explain and a quotable authority figure to back it up.
Tufte would most likely suggest that the “corrected” chart is still a visual lie. By restricting the left axis between 8 and 10 percent it suggests a more significant change than what really occurred.
Sure, the chart does look funny, but on the other hand, the honest chart can mislead too. While the honest chart looks like a drop in unemployment, the associated news was it didn’t represent so many people going back to work. Many people, especially women, stopped looking for work. Also, many of the new jobs are retail, so they will probably end after the holiday shopping season.
The headline of the article you linked to is “Hiring up, unemployment down”. So no, it’s not simply that more people stopped looking. And these figures are ADJUSTED for seasonal hiring, meaning that it has nothing whatsoever to do with the holiday shopping season. This is the lowest unemployment figure in the last two YEARS, by the way. That includes November and December of those previous two years.
Given the window that they are trying to stuff this into, whatever else is wrong with the graphic it is inevitably going to suffer from the data not being banked. It suffers more from lack of context. The range for the year-to-date is only 0.5% and, given the nature of the estimates, any number in the series could easily overlap any other number if there were any way of going out and actually measuring it. But even that isn’t the biggest shortcoming–history did not begin on January 1 and won’t end on December 31. We got where we are today through a process of historical development and you need to see the century-to-date data to put the month to month changes into perspective. Something like http://media.richard-careaga.com/img/unemp.png
Linearity not Fox’s strong suit http://imageshack.us/photo/my-images/809/foxlinear.png/
Hey! Both parties in the duopoloy suck–just like the data in this chart.
One instance of a shitty job on one piece of graph data does not an entire right-wing conspiracy make. I’d venture 90% of Americans can’t make that graph, much less read it…
90% of Americans aren’t employed by major news organizations to make graphs.
This is an important point: *I* would be likely to make some of the errors pointed out my the smart commenters here, but I am not paid to know how to visualize data for an influential, national media outlet.
With 2 seconds of review, however, one can also discern the angle of decline from January to February should be the same as the angle from Feb to March, but it isn’t. (Corrected in the second chart)
As lefty as I am, this chart isn’t evidence of right wing conspiracy, it’s evidence of ghastly quality control and lack of critical thinking skills. I try not to assume conspiracy when stupidity will do.
Could be a good time for FOX to use a test for statistical significance to help the general public…wait, nevermind.
They had the correct number there, didn’t they? It’s obviously an honest mistake. Or are you really so blinded by your bias against Fox News that you think they purposely made it look like it didn’t change, hoping their Neanderthal audience wouldn’t notice the number hanging over that point?
How is that “obvious”?
Yes, they put the correct numbers up, but the shape of the graph is clearly meant to make the drop in unemployment look smaller than it actually was. Given Fox’s long history of deception in both screen graphics and in spoken reporting, what possible reason is there to give them the benefit of the doubt on this one?
And as for your sarcastic suggestion that maybe they were hoping their audience would be too stupid to notice the numbers? That wouldn’t be a bad bet on their part, given that a recent survey showed that people who get their news from Fox are literally less informed than people who don’t have ANY news source.
Another possibility: What was being discussed while this was displayed? Did someone take this chart out of context to make Fox look bad? If the anchor was discussing the real unemployment as it relates to the size of the workforce and other factors and suggesting that the line would look more like what is shown if the numbers were reported honestly then it is the Fox detractors who are being dishonest. This would not be surprising since the source is credited as Media Matters. They’re not exactly an unbiased source of information themselves.
> you think they purposely made it look like it didn’t change, hoping their Neanderthal audience wouldn’t notice the number hanging over that point?
Bob I think that’s EXACTLY what they did.
It’s easy to fall in to R/D discussions. But please have a look at REAL unemployment as measured by the US up until the Clinton administration and you’ll see **23%** unemployment. http://www.shadowstats.com/alternate_data/unemployment-charts
What is the basis for regarding the scale on the vertical axis as “correct”?
This all assumes that 8.6% is an accurate figure, which is bs. It’s 8.6% if we don’t count all the people that dropped off the unemployment rolls for being on it too long and have quit looking for a job altogether, along with another of other factors that they use to manipulate the data to make it look better than it is. More people are unemployed, yet the number goes down. How does that even make sense? Only in the world of political spin.
The chart is still inaccurate, even if you disagree with the 8.6%. The values don’t fit the vertical scale. That’s undisputed. This analysis doesn’t assume that any of the values are correct — just that the chart doesn’t make sense.
You’re entitled to believe that there’s a better measure of unemployment than what the BLS publishes, but I think the story is still that the seasonally-adjusted employment situation improved in November — it’s not just an accounting trick.
Saif — the Fox News chart is still inaccurate, but the Media Matters chart that is “what the chart should like”, according to Flowing Data, still relies upon a completely arbitrary vertical scale. I don’t get it: Flowing Data surely knows better.
“Fox News – Fair and Balanced.” Meaning they need to artificially overweight the far Right (in their graphs) to maintain a balance with the Left.
I love how 8.6% is higher than 8.8%. Whatever, right? Just trust Fox on this one: “Things are staying bad or getting worse, so vote Republican.”
Somehow I’m thinking of that excellent book “How to Lie With Statistics”, whenever I see a chart with the bottom eighty percent chopped off… :)
I think both charts are actually misleading! By setting the y axis limits to start at a non zero value, it appears there is a big difference month to month. If the y axis minimum value was zero and the graph scaled appropriately, it would look almost like a flat line!
When you’re dealing with national-level numbers, a small percentage change means hundreds of thousands of people affected.
Nathan, by saying this, you implicitly acknowledge that your preferred graph, just like the Fox one, is slanted to make an emotional and political point. I don’t say that’s completely wrong — vertical scales are inherently arbitrary–but you of all people should know better than to imply that the vertical scale on the Media Matters is in some sense “correct.” To be frank, I am disappointed in you for that. Hundreds of thousands of people are still a small percentage of the labor force. If “flowing data” means anything, it should mean letting the data speak. If you want to show the individual tragedy of hundreds of thousands of still unemployed people, or the individual good news for hundreds of thousands of people who are re-employed, show that. There are a dozen interesting ways you could approach that with data that are appopriate for the job. Don’t take a blunt instrument (BLS data) and try to fashion it into something that can support a highly arguable and complex argument about the level esponsibility of any Administration for month-to-month economic growth.
@Fred – What I’m saying is that a small change on such a large scale is significant, and it’s why the media covers changes in tenths of a percentage point — whether it’s up or down — so closely. The context fits the form. BLS does this, too.
And in all fairness, I made the above graph for an easy comparison to the original. In a vacuum, I would’ve used monthly changes, but now we’re getting away from the point of this post, 8.6 ≠ 9.0.
Fair enough, that’s a reasonable rationale for why the media covers it that way. I read your post as a result of a link from a friend who was arguing that Obama has been responsible for some improvement in the economy. That’s why I was making my point that 8.6 ~ 9.0. I do agree that 8.6 ne 9.0.
The graphic arts people for the New York Times, etc. are well known — does anyone know anything about the graphic arts folks at Fox?
Wishful thinking? No. Honest mistake? No.
I think it was their intention to be deceptive. I think it was their intention to use this chart to mis-inform their viewers, and to misrepresent the truth.
I say it’s active deceit.
Every time I see a Fox News presentation (or hear one of their anchors speak), I think of Hanlon: Never attribute to malice that which can be adequately explained by stupidity.
This is pretty funny. I’m guessing what happened is they had Oct filled in to the chart and changed the label without updating the data. I could imagine how the conspiracy-minded might think otherwise and Jon Stewart could make a good 60 seconds of comedy out of this picture. But, if they really wanted to make the president look bad they could show U6 or U6* instead of U3 headline, or start the chart back at the month before his inauguration when the rate was 7.3%.
I’m going to have to go with intentionally deceptive on this one because seriously – how hard is it to make a line graph? Anybody with Excel can produce an accurate line graph given the same data that FOX uses. I can’t think of any way you could accidentally make a graph like FOX’s that fucks up the difference between 9.0 and 8.6 using any halfway respectable spreadsheet program. The only way is to intentionally jiggle with the numbers and the line graph or mislabel the data points.
You’re looking at it wrong…
Which is what Nathan and I discussed below beginning at 5:07 pm.
actually, the proposed chart correction is still wrong. the y-axis is misleading in that it does not include the whole range, starting from 0. If it did, then the drop in the unemployment would not be nearly as dramatic as many who have posted hear would like. I wonder if you folks were as critical about all the misleading charts presented in Gore’s movie.
Thank you very much. I found this first on Boing Boing and on Joe My God and have reblogged it at http://pedanticpoliticalponderings.blogspot.com/2011/12/repudiation-fox-news-fake-graph.html Your correct graph is particularly appreciated.
This, in my view, was intentionally deceptive.
“Making an evidence presentation is a moral act as well as an intellectual activity. To maintain standards of quality, relevance, and integrity for evidence, consumers of presentations should insist that presenters be held intellectually and ethically responsible for what they show and tell. Thus consuming a presentation is also an intellectual and a moral activity.”
Edward Tufte, Beautiful Evidence
I might get blasted for quoting Tufte for the sake of quoting Tufte, but this was the first thing that came to mind when I saw this chart. Flagrant and irresponsible, regardless of one’s political views.
I’ll let it slide this time.
I appreciate your mercy Nathan :)
Media Matters has a worst reputation for distorting information. Its not even a news organization its a Public Relations firm run by radical socialist. Unfortunately because of wide spread water fluoridation and use of antidepressants we have been infected by a population of drooling zombies with no critical thinking skills who call themselves progressive liberals. Obama is a pathetic sock puppet that is controlled by the authoritarian’s that program the fluoride heads through television.
Water fluoridation? That’s why I stick with rain water or pure grain alcohol.