With the relevance of US polling datasets now largely fading until 2022, investors are pivoting to legislative and regulatory forecast datasets, which they hope can help them assess the effectiveness and impact of a potentially divided US government.

Polling data once again performed less well than users hoped this year, with many pollsters incorrectly predicting a “blue wave” across the executive and legislative branches in the run-up to the election. While the outcomes of many elections are still being decided, it’s now obvious that that wave never materialised. Instead, the presidential election between Donald Trump and Joe Biden will be decided by razor-thin margins across several key swing states, including Arizona, Nevada, Pennsylvania, North Carolina and Georgia, and control of the Senate may come down to a January runoff election in Georgia.

Despite many vendors with incorrect predictions, some polling data providers told clients before the election that they expected results to be closer than what was being reported.

RIWI, a provider of public opinion surveys, released a report to its clients on Monday, November 2 indicating that it expected Trump to be re-elected. Its models showed Trump with 259 electoral votes versus Biden’s 208. It noted that the candidates were statistically tied on 71 electoral votes, but that Trump had a 9% probability of gaining at least the additional 11 votes that would ensure victory.

The data provider also predicted that Republican Senators were expected to win three of the eight key Senate rates, while Democrats would win three and two were tied. At the time of this article’s publication, several of these races were too close to call, but the report correctly predicted victories for Joni Ernst (R-IA), John Hickenlooper (D-CO), Tina Smith (D-MN). It incorrectly predicted the outcome of the Senate race in Maine, where Susan Collins (R-ME) defeated Sara Gideon (D-ME) handily.

The shift toward investors’ greater focus on legislative and regulatory datasets marks a trend that was brewing even before the election — after the polling errors in 2016, many investors took a longer-term view of the 2020 election rather than trading around the event itself, according to Daryl Smith, Neudata’s head of research.

Photo by Brianna Santellan on Unsplash