Why are Housing Sector Statistics Seasonally Adjusted?

As sure as foliage changes color in the fall and snow falls in the winter, economics statistics experience predictable changes during the year. To take these predictable changes into account, statistics are often seasonally adjusted. In seasonally adjusting economics statistics, analysts are better able to compare from one period to the next and better able to identify trends outside the norm.

Major economic indicators such as gross domestic product (GDP) are commonly reported as seasonally adjusted. Housing statistics commonly reported as seasonally adjusted to name a few, include housing starts, existing home sales, and residential construction employment.

Chart 1 below presents monthly single-family housing starts not seasonally adjusted. The chart highlights the up and down movement of the data series.

Chart1a

A similar effect is found the not seasonally adjusted monthly existing home sales data provided in chart 2 below. The data series partially obscures the long-run trends in existing home sales due to seasonal peaks and valleys.

 

Chart2a

The effect of seasonal changes is also clearly seen in residential construction employment. The chart below presents seasonally adjusted and not seasonally adjusted monthly employment statistics in the residential building construction sector.

ChartEmp

By presenting the statistics side-by-side the predictable movement of residential building construction employment that is not seasonally adjusted is clearly seen. January employment, for example, is typically lower than May employment in the same year. The seasonally adjusted series shows a smooth u-shaped curve whereby employment in residential building fell from January 2009 but has steadily recovered.

Seasons are not the only thing that change in a predictable manner. This post helps explain the rationale and benefits of seasonally adjusting selected series. Economics and housing statistics follow predictable patterns, which should be taken into account when analyzing time series data. Of course, the adjustments cannot take into account an unusual weather or seasonal event that is not typical such as the slowdown in construction that occurred earlier in 2014.



Tags: ,

Leave a Reply

Your email address will not be published. Required fields are marked *