You know what is inefficient? The US census. In the modern age, why are we trying to print out a form for everyone in the United States to fill out, put it in an envelope, slap a stamp on it and then send it to every single address? And then, as a backup measure, send someone driving around door to door to make sure people fill out the form (often carrying another copy with them), only to have huge numbers of people still not participate? Why on earth would we spend so many resources in paper, people, time, gas, and money when this kind of thing could be done in other ways?
What’s more, the data quality is remarkably low. How low? So low that the Census Bureau felt it necessary to point out that there is “little” evidence of falsified data in the census. They say only about 0.4% of respondents likely falsified data. One wonders how that gets verified or what is meant by falsified data. The word “falsified” implies intent. Yet, how does that account for unintentional inaccuracies? Or people filling out their forms either just before or after they moved?
Regardless, the Census Bureau was concerned enough about data inaccuracy that they did follow up interviews with over 300,000 households. That’s a pretty big response requiring yet more resources, all in the hopes of achieving greater accuracy.
Why does accuracy in the census matter so much? The biggest reason is that it directly affects representation in the state and federal legislatures. The amount of seats in the House of Representatives is directly affected by population. Population numbers, demographics and other data are also important to the distribution of 1.5 trillion dollars in federal funds as well. Or to put it another way, how your tax dollars are distributed and who makes those decisions are directly affected by the accuracy of the census data. Needless to say, it’s important to find the best, most accurate methods of gathering that data possible.
So, what reasons do we have to assume that the data is inaccurate beyond mere accusations? We know that they will sometimes rely on data from landlords, friends and family members if some people don’t respond directly. Some census agents have even been directed to make guesses based on the number of cars and bikes out in the driveway, or even by looking through people’s windows. There are so many problems with that, it’s hard to even know where to begin. A family could be on vacation so no one is around for a while, or they could have friends over for dinner, meaning there are extra cars parked out front. And are we seriously okay with the idea of government employees lurking around our yards peeking into the kitchen windows. It doesn’t take a lot of imagination to figure out all sorts of ways that could go wrong.
Sadly, this is often the case with data gathering overall. The census is just the archaic dinosaur version of data skimming, cookies, and selling your data to third parties without your consent. There are inaccuracies, falsifications, guesses and deceptions everywhere that some motive other than accurately representing people is present. Whether those other motives are profit or getting your manager off your back, it all leads to poor data quality.
What if the Census Bureau took a different approach? Rather than spending all sorts of money and effort to get people to take the time to fill out a form and send it back when there is no immediate reward for doing so they worked with TARTLE? They could offer a financial incentive to people to respond to all the same questions from their phone and get a financial reward in the process? It would be faster, more efficient, and almost certainly more accurate, especially if people could choose to send pre-existing data packets that already reflect exactly the kind of information the Census Bureau is after. By making use of the TARTLE data marketplace, the government would get a better understanding of the population and better represent it in Congress and in funding.
What’s your data worth?
