What predictions matter, Will I Be Successful? Matthew LoPresti

Personally, on some level or another we all want an answer to the question “Will I be successful?” Since success is a hard term in and of itself to define I will rephrase the question to the somewhat easier to predict “Will I be monetarily successful?” This question, while only about a single individual, requires us to draw data from many sources across large populations to provide a complete picture from which to make our assessment.

This problem calls for large scale statistical analysis using factors like race, education, family background and income, geographic location, and some measure of personal aptitude. To acquire this data you would need a government level census effort sustained across many years to track results and refine your model. The U.S. has had a long history of collecting census data dating back to the first one in 1790. In the modern day, socio-economic information is gathered by the U.S. Census Bureau's American Community Survey.

This methodology, while thorough, has its limitations. The models often hold true across large populations of like individuals when predicting medians but the variability of any one person’s circumstances can make the predictions unreliable on an individual level. This move towards the mass collection of data and its use in predicting human behavior and outcomes has been a staple of the modern internet era. As more and more of our lives move online, companies may be increasingly able to predict your financial and social situations as well or better than you yourself can. So while we all may want to know what our own financial futures hold we will increasingly have to weigh the cost as we narrow down the models.