Democrats took control of the House and Senate in the 2026 midterms despite higher Republican ad spending, a result that puts a record campaign outlay in focus, CNBC reported.
Political advertising totaled $12.06 billion over the 2025-26 cycle, more than $460 million above AdImpact’s June estimate. AdImpact said this was the first midterm cycle in which ad spending topped that of a presidential cycle.
Republicans spent $5.51 billion on political ads, compared with $5.11 billion for Democrats. Republican spending rose 27% from 2024, while Democratic spending fell 4%. AdImpact said it was the first time in more than a decade that Republicans had spent more on ads than Democrats.
CNBC’s analysis of Federal Election Commission records found that 14 of the 20 largest donor groups supported Republican candidates. SpaceX and Tesla CEO Elon Musk was the cycle’s largest contributor, directing more than $366 million to campaigns, CNBC reported.
Fundraisers told CNBC that Democratic donor hesitancy reflected Kamala Harris’ 2024 defeat, concerns about democratic socialism and distrust of party leadership. In that election, President Donald Trump beat former Vice President Harris even though Democrats spent more on advertising, CNBC reported. Democrats also flipped seats in Texas and Kansas, according to CNBC.
Broadcast television accounted for 49% of ad spending, or $5.85 billion. Connected TV brought in $2.66 billion and digital advertising $1.91 billion; AdImpact said both categories grew by double digits compared with the 2024 presidential cycle. Cable spending fell 7% and radio spending 13% from 2024, according to AdImpact data cited by CNBC.
Industry experts told CNBC that broadcast remains useful for reaching local markets and people likely to vote, despite falling viewership and competition from streaming. The Associated Press reported that Ken Paxton, a Republican in Texas, bought ads during national sporting events as a potentially less expensive way to reach voters statewide than purchasing airtime market by market.



