Reporting massively large numbers

By Felix Salmon
May 11, 2009

What are we meant to do when we read a story like this one, full to bursting with unimaginably large numbers? The 2009 deficit is now $1.84 trillion, up $89 billion from $1.75 trillion! That’s 12.9% of GDP! The spending plan for 2010 isn’t $3.55 trillion any more, it’s now $3.59 trillion! And so on and so forth.

I’m not picking on Reuters here at all — the team has actually done a spectacularly good job of trying to present these numbers in as many different ways as possible in order to give an idea of how big they are. But the problem is that the job is basically impossible because the overwhelming majority of human brains simply can’t comprehend the sheer magnitude of something like $1 trillion.

One thing which might help would be a cost-per-household measure. (As well as a hyperlink to the primary source.) In 2009, the figures now have total tax revenues of $2.157 trillion and total expenditures of $3,998 trillion, for a total deficit of $1.841 trillion. In real money, assuming 114 million households in the US, that means the average household will pay about $19,000 in taxes this year, but that the government will spend about $35,000 per household; the difference of $16,000 per household will have to be put on the national credit card.

Obviously averages conceal as much as they reveal, but at least this kind of exercise brings the numbers into the realms of the comprehensible. Once numbers go over $100 million or so, they pretty much cease to have any meaning for most of us, except as numerical exercises. It’s often helpful to bring them down to the kind of dollar figures that people can relate to.

More From Felix Salmon
Post Felix
The Piketty pessimist
The most expensive lottery ticket in the world
The problems of HFT, Joe Stiglitz edition
Private equity math, Nuveen edition
Five explanations for Greece’s bond yield
Comments
14 comments so far

Averages in skewed distributions can be very very misleading. How about median values for the figures you cite? Are they available?

Thanks for your fine blog.

Posted by Cliff | Report as abusive

Yeah,, right!! Obama owned this when he voted for the first stimulus package with Bush. He then rushed another stimulus package through with the help of 3 Senator’s help.

I’m still not sure the Toxic Asset Relief Program has helped with Toxic Asset yet.

Now we get to pay for this 3.4 trillion dollar budget.

And yet Govt. bond prices are up? Who is going to buy these bonds?

Posted by tomrus | Report as abusive

“…average household will pay…”
extremely misleading, first households aren’t the only ones to pay taxes and second, for their share, the distribution is extremely skewed, 40% is paid by 1% of households.
Reuters shouls stick to give the raw data ( + links to original is a good idea), the fact that people can’t comprehend the numbers is irrelevant, this-is-a-financial-wire.

I guess the alternative is between raw but incomprehensible numbers and comprehensible but misleading numbers… Thanks, but I will pick the former. I am also skeptical about the use of the “credit card” metaphor — as if public spending could be equated with household consumption, when in reality, most spending can be thought of as an investment susceptible of generating flows of income in the future: it may not make sense to put this flat screen TV on your credit card, but taking a loan to go to college or start a new business is not necessarily irresponsible. The “per household” average is also quite misleading since public spending generates positive externalities: spending levels that would seem unreasonable for the individual household can be optimal for a nation as a whole.

Posted by p | Report as abusive

I would focus on comparatives (both across time and space) and stop highlighting the nominal data.

The 1.8 trillion deficit is meaningless. We need to put it in context. Describing it as % of GDP is a step forward (but comaprison with households is not). But there are better measures. Government doesn’t pay its bills from GDP, it pays them from taxes. So a better number wold be the deficit, or debt, as % of all revenues.

Next you want to comapre across time. How does the current deficit compare with other periodsin US history? And then across space. How do other countries compare?

That wold be useful information.

re: “the difference of $16,000 per household will have to be put on the national credit card”

But the “credit card” currently has a near-zero interest rate. (Though, it would be interesting to know how much that could change; much is short-term versus long-term?)

Posted by Brian Slesinsky | Report as abusive

Thanks for this post. Govt spending has been > 1 Trillion for over 20 years. Repeating the numbers over and over serves to confuse, while some measure of journalism must be to inform. Describing numbers in a way people can understand is more towards the end of informing, than keeping them in a form they don’t understand.

If you, and all your descendents going back to Jesus’ time spent a million dollars per day your family dynasty would still have yet to go through a trillion dollars.
That honor would go to your great-great-great-great-great-great-grea t-great grandchild, sometime in 2710 (ish).

Posted by Dave | Report as abusive

Don’t forget: 1.3 of the 1.8 was already spent by Bush before Obama came into office. Republicans, take a good look in the mirror: That money WAS ALREADY SPENT.

Also, when you look at Debt-to-GDP, never forget that all Republican presidents end their term with a higher ratio than they started with.. and all Democrats end their term with a lower ratio than they started with. Bush2 followed the trend. And anyone still believes Republicans have ever been fiscally conservative?

Posted by Unsympathetic | Report as abusive

Budget per US household is a good idea as this brings the numbers down to more manageable levels, but that also brings in a variable (the number of households). Tax revenue also come from corporations, so that might be a figure to divide by (or some combination of households and corporations). If you want to bring the figures into manageable territory the dividing factor should make sense. If you divided by the number of households with a microwave oven or broadband internet, then you might even find the figures go down year on year…

Posted by Nic Fulton | Report as abusive

I’m not sure I agree with the overall concept. I’ll agree that numbers of high magnitude are routinely discussed in the Media, that’s undeniable.

The position that humans cannot comprehend what these numbers mean, is close to ludicrous.

A decent education, including a thorough grounding in mathematics makes digesting numbers less arduous. The reader with requisite analytical skills can do their own comparisons.

Good journalism helps of course. A clever writer can provide fast points of reference to allow the reader to quickly walk through a statistically dense piece.

The mistake so common in society today is to dumb down facts to somehow make them more comprehensible to Joe Sixpack. That will never work, because Joe will get a blurry idea of “what it all means” and then become his neighborhood bar instant expert on all matters mathematical, financial and so on.

Journalists have a never ending duty of synthesizing and interpreting information in an understandable manner, but that shouldn’t make Journalists substitutes for teachers.

Joe Sixpack should have gotten a sufficiently good education to permit him to understand the numbers. The education itself was free, let’s remember. If he didn’t care to, couldn’t be bothered to, or for other self indulgent or lazy reasons failed to, then the journalist has no responsibility to provide assistance for this meathead we’re talking about, to better understand what he’s stumbling though on the front page of his local newspaper.

Living through changing times is (and should be) difficult, but MUST we perpetually provide the ‘dumb crutch’? The answer is; no, we shouldn’t. We have no moral legal or ethical duty to do so, only some vague notion that large numbers of people “don’t get it”.

If ‘the system’ worked however, that would no longer be a concern for politicians, writers and journalists alike. They could get on with the message, and spend less time educating their readership/audience/constituents.

Personally, I have no patience for reinventing the wheel when it comes to public comprehension of complex datasets. That’s what education is for, and it’s time we got back to basics.

The piece makes a good point, but I don’t like the “overwhelming majority of human brains” part. It makes it sound too much like genetics and fatalism. Using the term “brain” makes it sound like an appliance does the job, rather than the educational system. I also find “overwhelming majority” to be a little overboard. Maybe it’s because I know so many people who are comfortable with numbers, and don’t have the least trouble with this kind of thing. I would substitute “the majority of Americans have not been trained to comprehend” large numbers. These can hardly be categorized as “complex data sets.”

Posted by Ron Evett | Report as abusive

With Peak Oil to hit the world economy in 5-to-10 years, the only way out of the huge Federal debt is to inflate it away. The government could default, but would then have to pay for oil with gold or threaten to stop all food exports. What havoc would that cause! I would plan for very high inflation starting within the next 5 years. People should also start thinking about how they will deal with gasoline rationing. There is simply no way to now avoid a reduction in our standard of living. You can’t spend more than you earn, & borrow the difference, forever.

Posted by Bill Simpson in Slidell | Report as abusive
Post Your Comment

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/