Rural means something different in California

8 September 2015 by Steve Blum
, , ,

California has been shut out of federal broadband grants for infrastructure projects in rural areas for the past several years. I was asked why don’t California’s wineries and farms throw lobbying dollars at the problem?

The wineries and farms don’t need to. California agriculture is characterised by large corporate holdings (albeit sometimes family controlled). The ag operations themselves can usually get sufficient connectivity, by building their own point to point microwave links and, occasionally, fiber connections.

The people who work on and around the ag operations, though, tend to live in concentrated communities in the 5,000 to 10,000 population range, and sometimes larger. That’s the situation in the Central and Salinas valleys, where they’ve been redlined by cable companies and left to wither by the telcos. USDA looks at that and doesn’t see a problem: the farms have at least 3 Mbps down/1 Mbps up or better, and farm workers who live in cities aren’t their problem.

At least, those communities are perceived as cities and treated accordingly. Even when the population is below the 20,000 mark that USDA usually sets as the threshold for rural status, the bulk of USDA staff – who are located in the south and midwest – see towns of a few dozen, or maybe a few hundred, people as their natural turf. USDA staffing tends to go on a county by county basis, with tiny southern and midwestern counties getting as much, or nearly as much, staff and resources as the huge counties in western states.

When the time comes to evaluate grant applications, the ones that resonate with them are the ones that are targeted to their core client base. It’s an institutional bias that needs to be fixed, but even if it can be – and I doubt it can be – it will be a long time coming. Communities and ISPs in California and elsewhere in the west need to crack the code and figure out how to make broadband infrastructure proposals fit within those existing biases.