Every once and a while a politician or political organizer or just a concerned citizen asks me what they can do to raise the minimum wage. I’ve been thinking and researching that question for a long time, so allow me to offer some thoughts/guidelines to contemplate, based both on the economic literature evaluating the impact of past increases and my own experiences over the decades. I’ll break this into two parts, the first on the economics and the second on the politics.

–Federal or sub-national: Despite the fact that increasing the minimum wage polls well among voters, enough conservative politicians at the national level are willing to do the bidding of the opposition lobby, headed by groups like the National Restaurant Association and the Chamber of Commerce. That’s prevented the White House and congressional Democrats from moving their bill to raise the national minimum from its current level of $7.25 to $10.10 over three years, at which point it would be indexed to inflation (the proposal also, and importantly, raises the sub-minimum for tipped workers).

I don’t see that changing anytime soon.

On the other hand, wage initiatives can still gain real traction at the state and city level. The “fight for $15” lives on, and according to the National Conference of State Legislators, “38 states introduced minimum wage bills during the 2014 session,” and “Connecticut, Delaware, Hawaii, MarylandMassachusetts, Michigan, MinnesotaRhode Island, Vermont, West Virginia and D.C. have enacted increases during the 2014 session…23 states and D.C. have minimum wages above the federal [level].”

Eventually the national wage floor will be increased, as it has been numerous times since its introduction in 1938. And not just when Democrats were in charge: the national minimum wage went up under both of the Bush presidencies. But for now, I’d focus most of my energy in this space on the sub-national level.

–The level: An increase must be high enough to reach the people who need it but not too high that it can’t be absorbed without significant distortions.

While opponents always claim that any increase in the minimum will lead to large numbers of job losses among affected workers, the evidence consistently belies the claim. History is quite clear on this: the vast majority of those affected by minimum wage increases benefit from them, though some analysis finds job losses at the margins.

For example, in their recent analysis of the impact of the proposed increase to $10.10, the Congressional Budget Office finds that 24.5 million workers would get a raise from the proposed increase, while employment would fall by 500,000, implying that 49 low-wage workers get a pay boost from the proposal for every one job lost. And other research suggests CBO may have high-balled the job loss number.

Instead, most of the increase is absorbed through the three p’s: prices, profits, and productivity. There’s a small price bump (economist Arin Dube finds that a 10 percent increase in the wage is associated with a 0.7 percent increase in prices). Paying for the mandated increase out of profits is an especially important mechanism given the sharp increase in profits relative to wages in recent years. And productivity effects are win-win, as higher pay is associated with significantly reducing two costly problems in low-wage sectors: high turnover and vacancy rates.

With that in mind, I can offer two separate guidelines for thinking about levels. First, my rough take of the history of minimum wage increases is that they typically capture up to 10 percent of the workforce in their “sweep,” meaning the share affected by the wage change (basically, the percent of the workforce earning between the old and new minimums). So policy makers can feel comfortable in terms of historical precedent when in that range (which, according to CBO, would include the proposal to go up to $10.10 in three annual steps).

Second, today in other advanced economies and earlier in our own history, the minimum wage was about half the median wage, and that relationship supported a more equitable alignment of wages in the bottom half of the wage scale. Today’s U.S. minimum is about 40 percent of the full-time median wage. Were the $10.10 proposal enacted right now, it would meet this benchmark, but remember, it has a three-year phase in, so that level is actually a bit weak by this metric.

It’s very useful to keep this “half-the-median” idea in mind when thinking about the more localized increases, like the highly publicized increase to $15 in Seattle. According to Professor Dube, after accounting for the phase in, the Seattle minimum will “likely be around 59 percent of the local median full-time wage…” which is high by historical standards but not off the reservation. This is also a reminder that the nice thing about “half-the-median” is that it can account for the high degree of local variation in wage levels.

— Adjusting for inflation: Finally, it is important to adjust the minimum wage for inflation. The fact that the federal minimum wage is not indexed was responsible for its nine year, 30 percent real decline during the Reagan years.

The inflation adjustment has obvious advantages in terms of not eroding the buying power of low-wage workers, but it’s also a plus for their employers, who gain some certainty as to their labor costs and are less likely to have to absorb large, intermittent increases.

It’s worth noting that 11 states currently index their minimum wages to inflation, and it’s worked well for them. In fact, there’s suggestive evidence that the only states where real wages actually went up in 2014 are the ones who raised their minimums. Notably, employment also did a little better in those states.

So, summing up: if you’re pitching a minimum wage increase, think sub-national for now, set the level as suggested above, and index for inflation.

Next installment: the politics of an increase.