Coherent Trends, Turning Points, and Forecasts for ACS Data
Tucker McElroy
KEY WORDS: Filtering, Frequency Domain, Nonstationary time series.
ABSTRACT
The American Community Survey (ACS) provides one-year (1y), three-year (3y), and five-
year (5y) multi-year estimates (MYEs) of various demographic and economic variables for each
\community," although for small communities the 1y and 3y may not be available. These survey
estimates are not truly measuring the same quantities, since they each cover different time spans.
We present a metric to measure the compatibility of different MYEs; for those that are deemed
to be sufficiently compatible, we describe methods for generating trends, turning points, and
forecasts of ACS data at 1y, 3y, and 5y intervals, in such a way that the different estimates can
be compared with one another. The filters utilized are non-model-based, require only a short
span of data, and are designed to preserve the appropriate linear characteristics of the time
series that are relevant for trends, turning points, and forecasts respectively. The basic method,
which only requires polynomial algebra, is outlined and applied on ACS data. The resulting
filters are analyzed in the frequency domain.
CITATION: