As we head to the midterm elections, the most important thing that any well-intentioned politician can do is prepare for the worst.  That doesn’t mean just imagining what might happen; it means playing games.  Real-life games that involve actual practical exercises to rehearse what an appropriate response would be in times of crisis.

When I was at the Department of Homeland Security, we often engaged in these sorts of Table Top Exercises (TTXs).  The objective was to create realistic scenarios of a crisis and play them out with officials making real-time decisions and seeing how well (or poorly) they worked.

These games were, candidly, kind of fun to play.  But they had a deeply serious purpose – to allow senior decision makers to understand concretely the nature of the problems they would face; the legal and policy restrictions that would bind them; the scope of the resources available to them; and, most importantly, the practical aspects of how quickly and effectively a response might be deployed.

At DHS we did TTXs to prepare for things like our hurricane response or how to assure the continuity of operations and government in the wake of another terrorist attack – and we always learned valuable lessons about what might or might not be possible. 

The same is just as true for state and local officials responsible for our elections.  The necessity of practicing in anticipation of a crisis is just as substantial (perhaps even more so) when democracy may be at stake.

To see the value of TTXs, consider one example (from many possibilities):  What can or should a Governor, or a Mayor do if ICE officers appear near polling places in a predominantly Democratic area of the city to “check the IDs” of voters under the guise of “making sure” that all voters are lawfully entitled to vote?

Naturally, the Governor and the Mayor might suspect an ulterior motive – ICE’s presence at polling places would have the impact of dissuading voters from exercising their franchise, with the inevitable asymmetric impact on pro-Democratic turnout.  How should State officials who wanted to mitigate that base motivation proceed?

The first thing they might do is turn to their Attorney General and ask about the law.  They would learn that Federal law (18 U.S.C. §592) makes it a crime for any armed federal agent to be stationed “at any place where a general or special election is held” or for any federal officer to order any agent to a polling location.  Meanwhile, another federal law (52 U.S.C. § 20511) makes it a federal crime to intimidate or attempt to intimidate any voter or anyone who is urging or aiding a person to vote.

But these laws are not self-executing.  Faced with an agency that has demonstrated (to put it politely) a willingness to stretch the boundaries of law, the Governor might ask the AG what he could do?

One possibility is going to court.  But just how effective would litigation or the threat of it be?  It might tell us, in practical terms, what the deterrent impact (if any) would be of threats by local district attorneys to prosecute the ICE agents for violations of state criminal laws.

A good TTX would also explore the possibility of real-time legal intervention on election day.  While courts have consistently ruled against much of the Administration’s efforts to interfere in local election operations, there is at least some reason to doubt whether a court-based response would be quick enough.  An exercise would, I expect, tell us that it would take at least a few hours before a court would enjoin Federal activity.   Going through the exercise would help us understand the scope of the possible delay and assess whether litigation alone would be enough to mitigate the harm.

Another value of a TTX would be exploring exactly how officials would respond on the ground.  Do we have clear lines of communication from, say, the local election officials to local law enforcement?  How would local law enforcement respond if called upon – what is their protocol?  For real-world examples, during the recent 2026 primary elections in San Antonio, Texas, and Simi Valley, California, local law enforcement agencies were called in because of reports of ICE agents near polling locations.

In both those cases, local police were responsive, and the ICE agents left when confronted.  But what would happen if the local police were not as responsive (perhaps because of their own political beliefs or resource constraints) or if ICE agents refused to leave?  How would a local sheriff’s office, say, be able to enforce its demands?  What if there is a gap and no state law clearly applies?

All of these are questions worth exploring in detail and they are best examined not in imagination or dry policy papers – but rather through practical exercises against real-time “opponents” who are creative in their opposition (what, for example, happens if ICE won’t identify itself?  If they are undercover?  If they aren’t armed?  Etc.?)  It may well turn out that the most important factor in responding to an ICE intervention is neither law nor policy, but the nature of the personal relationship between, say, a Mayor seeking to protect a polling place and her Chief of Police – and a good TTX would help them coordinate their response.

And these are just the complexities of responding to a single scenario.  It is easy to imagine a host of other variations, each of which is deserving of exploration: 

• What if instead of ICE it’s the FBI?  Does that make a difference? 

• What if it’s the military?

• What if the Governor wants to deploy the National Guard in response instead of local law enforcement under the Mayor?

• How would a Mayor and Governor coordinate?  What if they are from different political parties?

Nor is there any magic to the scenario of an ICE ID check.  While we are at it, we should consider some even worse possibilities.  How would the Secretary of State of Georgia and the Mayor of Atlanta (to take one hypothetical) respond if instead it was the FBI arriving at the ballot counting facility with a facially valid warrant to seize the ballots?  Or the DEA conducting an area-wide cordon and vehicle checkpoint because of a “report” of drug violence adjacent to the polling station? 

The parade of horribles is as long and detailed as our imagination can make it.  Which is precisely why playing these horrible games in advance is essential.  Imagine the worst – and then play the play for victory.  Saving democracy demands it.