How to compare wages in two states fairly

“Nurses earn more in California than in Texas” is a sentence the state pages can support or refute. Whether the comparison is fair depends on five things being held equal, and on being clear about the one thing this site does not hold equal.

Same occupation

Check the SOC code, not the title. Registered nurses (29-1141) and nurse practitioners (29-1171) are different codes with different distributions. A comparison that takes one state’s figure for a broad occupation and another’s for a detailed one inside it compares a mix with a part of the mix; see how BLS defines occupations.

Same statistic

Median with median, 90th percentile with 90th percentile. A state’s mean against another state’s median overstates the first; see what the mean wage hides. Hourly with hourly, or annual with annual, and if one state’s row is annual-only for the occupation, the other must be read the same way.

Same reference month

Both figures should come from the same OEWS release. Every state page on this site prints the reference month, and both states in a comparison on this site always share it, because the site holds one release at a time. Figures copied from elsewhere may not.

Employment in view

A state with 400 workers in an occupation has a noisier estimate than one with 40,000. BLS publishes a relative standard error for every cell; the state tables show employment so a reader can see how much weight a figure carries. A difference of a dollar an hour between two small-sample estimates may be nothing.

The floor

For occupations near the bottom of the distribution, the comparison is partly a comparison of minimum wage laws. If one state’s floor is $7.25 and the other’s is $16.00, the 10th percentile for retail or food service workers will differ by roughly that amount regardless of anything else. Read the state’s minimum-wage page next to its occupation table.

What is deliberately not adjusted: cost of living

A higher wage in a more expensive state buys less. Readers often want the wage divided by a price index, and this site does not do it. The reasons are specific.

There is no official index of price levels by state that BLS produces; the Consumer Price Index measures change over time, not differences between places. The Bureau of Economic Analysis publishes regional price parities, which do measure relative price levels across states and metros, but they are built for aggregate personal income, weight housing heavily, and move a state’s rank when the method changes. Dividing an occupation’s median by a regional price parity gives a number that no employer pays and no worker receives, and the result would appear beside the published figure with the same authority.

The site therefore shows the nominal wage, labels it as such, and leaves the adjustment to tools designed to take a reader’s own circumstances: the sister site FastCalc has take-home and paycheck calculators that start from an hourly or annual figure and apply the state’s withholding. This site publishes the inputs and computes nothing.

A fair comparison, in practice

Open both state pages. Find the occupation in each table. Confirm the code matches, read the median and the 25th-to-75th range, note the employment and the reference month, and look at each state’s minimum wage if the occupation is low-paid. State the result as “the median for X was $A in state P and $B in state Q in May of year Y,” which is what the data support, rather than as “workers in P are better off,” which they do not.

Neighboring states are often the most useful comparison, because labor markets cross their borders; the metro tables on each state page show where the two states’ metros overlap.