POLK COUNTY BROADBAND PLAN
PROJECTION OF FUTURE DEMAND
Although some methodologies for projecting broadband demand have been developed in the past,37 no existing
methodology was considered appropriate for applying to the Broadband Polk planning effort to project future
demand. This was due in part to the unique characteristics of Polk County, in part to the limitations of existing
methodologies (for example, some methodologies were designed for use in other countries, and some are based
on outdated broadband technology such as DSL), and in part to the limitations of available data. Therefore, after
a thorough literature review of broadband demand modeling methodology was conducted, an innovative spatial
model was produced to project future broadband demand in Polk County.
This model projects broadband demand in the year 2020 by producing a weighted broadband demand score
for all of Polk County, Florida. The weighted score is a number between 0 and 5, where 0 indicates the lowest
broadband demand and 5 indicates the highest demand. The model inputs are Future Land Use, projections of
population and employment, and estimates of current broadband penetration.
The smallest geographic units of measurement are land parcels. The model could also be aggregated to the level
of census blocks, block groups, or tracts.
The model equation is as follows:
B=E+P+F+C
Where:
B is the broadband demand score (a number between 0 and 5).
E (see Figure 55) is the percentile rank of projected employment per square mile in 2020 for each TAZ. There are
621 TAZs in Polk County. This model ranks them according to their projected employment density in 2020, then
assigns a percentile value (i.e. a number between 0 and 1) to each TAZ. The TAZs projected to have the highest
employment density are in the 99th percentile, and thus receive a score of 0.99.
Figure 55. Percentile Rank of Projected 2020 Employment Per Square Mile (E)
37 For example: Carlo Hjelkrem, Kjell Stordahl, and Johannes Bøe. “Forecasting residential broadband demand with limited information – A longterm supply and demand model.” Telektronikk 4.2004. http://www.telektronikk.com/volumes/pdf/4.2004/Page_043-049.pdf
Kjell Stordahl and Lars Rand. “Long term forecasts for broadband demand.” Telektronikk 2/3.1999. http://www.telektronikk.com/volumes/
pdf/2_3.1999/Page_034-044.pdf
73