“Stand your ground” laws are very misunderstood and heavily politicized.

And from a data perspective, they don’t do much at all, though they appear to be a net positive.

Take-aways

• “No duty to retreat” states saw mild declines in both total homicide and gun homicide rates.
• The majority of “no duty to retreat” states saw declines, where the majority of other states saw increases.

Can You Stand It?

The name itself, “stand your ground,” is a heroic meme and a bit of a misnomer. However, the non-politicized term “no duty to retreat law” is both cumbersome and provides neither the pro- or anti-gun factions with a phrase to inspire their followers… or fill them with dread. Indeed, “stand your ground” (the pro-gun phrase) evokes a “cowboy American” posture while “shoot first law” (the anti-gun phrase) evokes endangering chaos.

Stripped of such fluffiness, “no duty to retreat” laws are two things simultaneously:

• An extension of the “castle doctrine”
• A limited immunity from prosecution for not evading an attacker

We’ll take those in reverse order. These laws say that you may use deadly force when you reasonably believe it to be necessary to defend against certain violent crimes. This is extended to saying you are under no duty (legal obligation) to retreat before using deadly force in self-defense, along with sundry conditions (e.g., you are in a place where you are lawfully allowed, you are not the instigator of conflict, etc.).

These laws are an extension of the ancient “castle doctrine,” which comes from the old adage that a person’s home is their castle. The castle doctrine allows you to assume that a person not allowed in your home (i.e., burglars, aggravated exes, etc.) can be assumed to intend you harm, and you are within your rights to kill them to protect yourself. “Stand your ground” laws are in effect castle doctrine brought onto the streets.

In this century, 20 states have passed such laws, so the question on today’s table is whether they did harm, did good, or did nothing at all.

Data Constipation

Fortunately, data is available for all but one of the states in question, but for two states we have reservations about data quality.

Montana, the 48^th least populated state by population density, has so few homicides – by guns or other means – we encountered perpetual “divide by zero” errors, and so we omitted them from the analysis.

But two other states showed a quirk. CDC data says that both South Dakota and New Hampshire had homicides but no gun homicides for the years 2000 through 2018. This is inconceivable, and the FBI crime stats for 2018 (as a spot check) show otherwise. Now, the CDC “suppresses” data when the numbers are too low, and we see very few gun homicides in these states (8 and 12 respectively for 2018, per the FBI). For now, we’ll allow the CDC to encode these as gun homicide rates of zero, but the CDC should provide an option for presenting the statistic even when data is in the statistical noise range.

That aside, we can look at both total homicides (murder by any means) and gun homicides. We would have liked to report on the effect “stand your ground” laws had on aggravated assaults, but the FBI neutered their data exploration tool, and other sources (NACJD) are missing entire years of data. Perhaps in a more data-rich future…

Spaghetti Charts on the Tables

click for a larger, sharable version

First, we looked to see the changes before and after passage of “stand your ground” laws against states that did not pass them. Since these laws were passed in waves, it was easy to cluster them together in groups (A, B, C, etc.) based on the year of the law becoming effective. We measured the slope of the line for before and after passage for each state and for the aggregate of states that did not enact such laws covering the same years. The green cells show where homicide and gun homicide rates were falling, red for when they were rising.

For all modes of homicide, rates fell in 58% of the states, and for gun homicides, in 63% of the states. Contrast that with the fact that the “static states” (states that did not pass such laws) all saw total homicide rates rise, though for gun homicides only half did. All other things being equal (which they are not), “stand your ground” laws are a net social positive.

However, rates were not evenly spread throughout all “stand your ground” states, though increases or decreases were generally nominal.

click for a larger, sharable version

click for a larger, sharable version

The simple, aggregated view tells us something interesting. On average, newly minted “stand your ground” states saw a net drop in gun homicides, with some minor undulations. Their cousin “static” states had an odd dip 2–4 years in. Looking at the data, we see two low-population density states – Alaska and Iowa – reporting CDC suppressed data in that span, which caused their average gun homicide rates to fall as low as 36% for the five-year pre-law averages. Hence, the dip is a data anomaly, as evidenced in part by how the gun homicide rates snap back in year +5.

As for all the states, this messy looking chart shows the difference between the “stand your ground” states and their sister “static” states. Aside from (a) single-year dips and surges and (b) one rising rogue state (Nevada), gun homicide rates were basically flat. On the chart, a rate of 1.0 means the “stand your ground” state was no different than the states that did not have such laws. The heavy line is the average of all the states.

Standing Still

At best, we could say that “stand your ground” laws have a mild net positive, with 58–63% of states having net declines in homicides and gun homicides respectively. But the margins are thin and the long-term changes are, by and large, flat.

The better summary is that, in aggregate, “stand your ground” laws don’t cause public harm and at least keep good actors – those committing acts of self-defense – out of trouble.