Thursday, December 10, 2015

One Versus Five Year Estimates: Measuring Regional Poverty

With the release of the 2014 Five Year Estimates by the Census Bureau, we now have the ability to compare changes between two five year periods: 2005-2009, and 2010-2014. For the region (Herkimer and Oneida Counties combined) we have always been able to compare the one year estimates. These have been available since 2005. So why would someone want to compare five year estimates then?

Both the one year and the five year estimates have certain advantages. If you are interested in the MOST CURRENT DATA, then you should look at the one year estimates. These are the most up-to-date data available. However, if you'd like the most accurate data, then the five year estimates are the better answer. They have lower margins of error due to a larger sampling taken over five years and provide the most precise measure of the data.

Let's compare the two on a single graph. Below is a chart showing the one year estimates (in blue) from 2005 to 2014 for our region for the percentage of people in poverty. As you can see, there is a slight decline in poverty prior to the Great Recession (from 2005 to 2007), and then a huge spike following that event (note the rise from 2007 to 2011).


In comparison, looking at the single year data may not give you the best idea of what has occurred. Because of the annual fluctuations, the overall trend gets a little lost - especially when you look at the last five years from 2010 to 2014. There are huge variations year to year, so how are you to get an accurate view of what is happening with poverty? This is where the five year estimates are helpful.

In looking at the 2009 Five Year Estimate and comparing it to the 2014 Five Year Estimate, you can see that there has been a substantial (indeed, a statistically significant) change between these two periods.In the last half of the last decade the poverty rate was at 14%; over the last five years the rate is now 16.4%.

PLEASE NOTE: THE FIVE YEAR ESTIMATES ARE NOT AVERAGES OF THE INDIVIDUAL FIVE YEARS. Rather, the data is collected over the five year and then measured as a whole.

In conclusion, both tools - the single year and the five year estimates - offer different advantages. Which one you prefer to use depends on what you are trying to measure !