WHAT: Form of the Data

You have now decided on the appropriate time frame and variables and now you must decide on which form of the data you want to use.  At this point there is no way I can provide you with a thorough treatment of this crucial issue, but I can briefly mention a few 'tricks of the trade'.  It may help as we go through these 'tricks' to keep in mind that in most cases we are simply attempting to place individual numbers in some perspective.  Any number by itself is meaningless at best, and at worst misleading.

In general there are two ways of providing some perspective. The first we would call cross-sectional perspective.  Is our income higher, our unemployment rate lower, our growth faster than others?  Perspective allows us to know how we are doing relative to others. The second would be time-series perspective.  How are we doing today relative to yesterday, to last month, last year?  This provides us with some historical perspective.

We will begin with one of the most important techniques, one that will allow us to adjust our data for the distorting effects of inflation. This is a must to make sense of both cross-section and time-series analysis when using data that are measured in currency. For example, what about the changing  welfare of American workers in the 1970s when average wages grew by 7 percent which was substantially higher than the 4 percent in both the 1960s and 1980s?  Or the welfare of Alaskans where average annual pay ranks first in the US, nearly 25 percent higher than Rhode Island? It turns out we can say very little based on these data until they are adjusted for prices / inflation.

In addition to adjusting data for prices, we also often need to adjust data for scale. For example, what about the safety of children in Rhode Island given that the number of reported cases of child abuse in Massachusetts outnumbers those in Rhode Island by nearly 4 to 1?  What about the fact that in 1994 there were approximately 60 times as many poor people living in California as there were living in Rhode Island? Is poverty a bigger problem in California? And what about death rates in Rhode Island that are about 12 percent above the national average.  Does this suggest that Rhode Islanders are at greater risk? It turns out that once again the numbers reveal little useful information, but in this case it is because size matters, something explored in the bench marking section.

We will now turn to a discussion of inflation adjustments and benchmarking.