The whole public debate about sequestration, cutting the deficit, and stimulating the economy is looking in the wrong directions. The broad solutions are simple: (1) the federal government should not spend as much as it does in relation to its income; (2) the government should target its fiscal policies in a way that makes the US economy expand. The details, of course, are where everybody bogs down.
I’m an engineer. I like to solve problems by looking at data and figuring out where to apply pressure to a system to get it to do what I want. Clearly, where I apply that pressure matters: it’s better to target the big-ticket items than the small fry. This approach means, simply, that cutting federal discretionary spending is almost completely irrelevant. Instead, proposals for managing federal spending should be looking at cutting things like the military. (This is one good thing about the sequestration plan: it forces the issue of cutting our ludicrous amount of military spending.) The politicians resistant to touching the military budget sometimes argue about the number of jobs involved – not just soldiers, but civilian contractors to the military. We wouldn’t want to hurt the economy by cutting military spending, right? Well, as it turns out, the military budget is not well-correlated with GDP growth. One reason for this result might be that, while the military is certainly interested in investments, infrastructure, advanced technologies, and new medicines – all things that can make for jobs and growth in the wider economy – the military also isn’t exactly interested in sharing those things with the civilian community through a commercialization process. It wants to invest in itself alone. This does not help make the economy grow. So, there is plenty of room for cuts in the military budget (and plenty of room to remain comfortably secure, too).
A well-crafted budget plan should also look at diverting spending towards programs that have the greatest positive impact on our economy and society. Yes, I’m talking about increasing some areas of federal spending as part of a deficit reduction solution. That’s because of multiplier effects – sometimes, the government can take actions that reverberate throughout the economy and generate positive benefits for everybody: jobs and wealth for citizens, increased tax revenue for governments. Win-win!
Increasing spending happens to be about the same in deficit and revenue terms as cutting taxes, but the multiplier effect from tax cuts isn’t going to be much to help. They provide some amount of economic growth, but there are a lot of studies that show that the effect is less pronounced than changes in government spending. Here is a good article outlining both sides of issue. In my opinion, the preponderance of evidence is that most economic growth for every dollar cut from federal taxes is lower than the economic growth from boosting spending. However, even the studies that don’t agree with me tend to show that most of these actions have multipliers up to about 1.6 – for every $1 of taxes cut or spending increases, GDP grows by $1.6. As long as this number is greater than 1, there’s a positive effect on the economy, but a 60% return on investment may take a while to have positive effects in society at large.
Fortunately, there are some slam-dunk areas where a little government investment goes a long way. One example is highway infrastructure investment: it may not be sexy, but it apparently carries a multiplier greater than two! This means the if the federal government cut $2 from the Pentagon budget, but invested $1 in the Eisenhower Interstate Highway System, then not only would the deficit shrink by $1 but the economy would grow by $2! (Plus, we would have bridges that don’t fall down.)
Even highway spending, though, isn’t as good as the government could do.
There’s this one government program that happens to provide a staggering return on investment, and is hugely popular with all demographics, but doesn’t really get a lot of federal budget love. It’s called the National Aeronautics and Space Administration. (I bet you were wondering when I would say something about space!)
For every $1 that the government spends on NASA, it spends about $200 on other things. But for every $1 the government spends on NASA, the economy grows by….well, a Freakonomics panel says that the economy grows by $8 – an 800% return on investment. A Rutgers University report posted on the Johnson Space Center web site puts the return at $7 (not just for NASA, but for research and development in general, as well). And here’s a link to a 2002 article that suggests that for $64 million of investment from the government through NASA, private companies received a “value-added benefit” of $1.5 billion, making a ratio of over 1 to 23. If a broker came to you offering an investment account with a historical 2300% rate of return, wouldn’t you take it? Purely as an engine of economic investment, without getting into any of the scientific, technological, or sociological benefits, NASA is a tremendous success!
Certainly, our national representatives should be engaged in a thoughtful and difficult discussion over what programs to reduce and which to expand. If they are smart about it, though, they should look at preserving – or even enhancing – those programs that benefit us the most. They should look at the data and target their actions.
Therefore, cut defense – I have enormous confidence that the Pentagon will successfully figure out how to prioritize. Cut some entitlements – there are certainly bloated programs out there. But fund infrastructure. Fund research and development. Fund the NSF, NOAA, NIST, DOE, and USGS. And fund NASA!
Quantitative Revolution
We’re going through an interesting sort of revolution in America. One after another, various disciplines are realizing (or, it’s coming out publicly that they have realized) that math is useful for stuff.
Wherever there is data available, a scientific, quantitative approach allows people to do two things. First, they can use existing data to develop a model which fits all the available observations. Next, they can in turn use the model to predict future behavior. And if people can make predictions, they can try to make decisions. Influence outcomes. Optimize certain results.
An obvious place for such an approach is the world of high finance, a discipline which is totally steeped in numbers and data – and completely focused on the very quantitative problem of maximizing a return and minimizing loss – but for a long time apparently ignored statistical modeling. People successfully applied statistical analysis, and ended up doing very well…but there was a backlash. Here’s an interview where a reporter complains that trying to optimize stock market gains somehow mis-values the stock market, at least according to his conception of value.
The Daily Show with Jon Stewart
Get More: Daily Show Full Episodes,Political Humor & Satire Blog,The Daily Show on Facebook
Geez. Those…those…physicists. They use models based on data of past performance, then try and predict future performance…and worst of all, they keep getting their predictions right!
(I want to note that if someone has a problem with the idea that these “quants” have privatized tremendous gains and socialized tremendous losses, that’s not a problem with their approach. It’s an issue with the goals of their models, and whether those goals are morally justified is a separate question from whether the approach works to satisfy the goals.)
We also have a ton of data available in the world of professional sports. Commentators make it their business to know – and inform viewers – whether or not this is the guy who gets on base with a ground-rule double on an overcast Tuesday more than any other player with an odd jersey number when the pitcher throws a 96-mile-an-hour fastball. In fact, this revolution I’m referring to might even be called the Moneyball effect. After all, that movie brought this idea forward in the popular consciousness.
Most recently – and certainly most dramatically – we have people who build statistical models on political poll data. Despite a constant media barrage insisting that the 2012 election was a dead-heat horse-race fifty-fifty hyphenated-adjective toss-up, these poll wonks stubbornly viewed their data scientifically, constructed careful algorithmic models, and predicted a much more certain, though far less entertaining, outcome. There was quite a backlash against these predictive models, at first, though the backlash seems to have been driven by either ideological preconceptions or a misunderstanding of the statistics: a poll showing two candidates with a 51-49% split doesn’t mean that the likelihood of each candidate winning is 51% or 49%. In true Hari Seldon-like fashion, the models aren’t predicting what single voters do or making decisions for us; but with an aggregate of people, they can make astonishingly good predictions. In many ways, this was the biggest story to come out of the 2012 American elections: scientific thinking and mathematical methods actually work!
This notion seems revolutionary, in each field it has touched so far. That appearance is what I find most surprising! Science has given humanity an entire body of knowledge. We can predict the behavior of quantum particles. We can determine whether there are planets orbiting other stars. We can forecast snowfall to within a few inches of accuracy a week in advance. We can find out what the feathers on a dinosaur look like. We can reconstruct Pangaea in a computer. And all the predictive mathematical models that allow scientists to do those things also give us cell phones, Angry Birds, medications, contact lenses, and all sorts of other goodies. Science isn’t just something that happens in isolated labs – it gets out into the world. And quantitative thinking isn’t magical wizardry – it is a tool that anyone with the will to apply themselves can learn.
This is a lesson that I hope we take to heart.