"Lights Out" and "One Second After" both make a lot of unwarranted assumptions about an EMP. They assume it works, while it was observed in some cases it was not observed in other cases. Nobody, not even the experts, are sure what an EMP device would look like and they don't know how to employ it if they had one.

The effect is not a sure thing. Many of the references used to support a claim of a catastrophic effect are quite old. Those books claim that modern electronics are always and predictably more vulnerable. Comparing mid-50s tube electronics to CMOS they clearly are.

Comparing more recent circuit and component designs, which are designed to avoid issues with voltage surges from any source, and optical systems which are entirely immune it is far less clear that an EMP would be more an issue now than in the 60s. Some quite credible studies claim that the vast majority of observed effects on test circuits in the 70s were temporary, requiring little more than resetting systems, or requiring replacement of simple components such as fuses.

Standard assumptions like vehicles with electronic ignitions failing is not so simple. Some tests on individual vehicles show they fail. Others not. As far as I can tell nobody has made a systematic study.

The doubt adds its own dynamic. Not knowing how effective, or ineffective any attack, an attack nobody knows how to do, an attack that has never been tried, may be makes it a high risk strategy. If you spend millions on a bomb would you use it on a potential light show and making our lights flicker?

Beware reading the survivalist literature and semi-factual alarmist sites too literally. They do not sell because they are reliable and accurate guides but rather because they are calculated to raise passions.