We're going to be telling a pretty complex story today, about the first time a city with more than 25,000 residents increased the minimum wage to $15.00 per hour in the United States.

Let's start with some basic background for the City of SeaTac, Washington and the way its $15 per hour minimum wage came to be....

SeaTac and Its $15 per Hour Minimum Wage

The city of SeaTac, Washington is located between the major cities of Seattle and Tacoma, where it is about a 30 minute commute from the downtown areas of both cities during rush hour and a nearly 50 minute commute from nearby Bellevue. The city has one main industry centered on one main location, Sea-Tac International Airport, where some 4,700 people toil in commercial aviation transportation-related businesses at the airport, accounting for more than a third of the estimated 13,500 people who work within the city limits of SeaTac, which itself is home to over 27,875 people.

In 2013, SeaTac made the news for something other than the airport and being the home to a prolific serial killer. On 17 May 2013, the first proposition measure in the U.S. proposing to raise the minimum wage to $15 per hour made the ballot in the city for that year's November elections.

The Prop 1 ballot measure was aimed at the city's most-captive industry: businesses supporting Sea-Tac International Airport, where the organized labor and city politician backers of the measure sought to increase the minimum wage that would be applied to people working at the airport and also the city's transportation and hospitality businesses that serve the airport.

On 6 November 2013, the nation's first $15 per hour minimum wage measure was approved by the city's voters and became the statutory minimum wage for workers in the City of SeaTac.

But not for all of the estimated 13,100 citizens of SeaTac who had jobs at that time. Several large business based at the airport quickly sued after the measure was passed and obtained relief from a Washington state judge, who ruled that the city's new $15 per hour minimum wage would not apply for the 4,700 workers at the airport itself.

The new minimum wage did however go into effect for the 8,400 people in SeaTac who did not work at Sea-Tac International Airport on 1 January 2014.

A year after SeaTac's higher minimum wage went into effect, local news reports described little impact to SeaTac's overall economy. KING 5 reported that some 1,500 jobs in SeaTac were directly impacted by the minimum wage at the time it went into effect, meaning that these jobs previously paid wages that were less than $15 per hour. KING 5 also estimated that 400 of these jobs, or 26% of the directly affected total, were held by individuals who lived in the city of SeaTac. The Puget Sound Business Journal reported that SeaTac Mayor Mia Gregerson was "not aware of any business closing because of Prop. 1", which was seconded by SeaTac's city manager Todd Cutts.

It would not be until 21 August 2015 that Washington's state supreme court ruled in the City of SeaTac's favor to extend its $15 per hour minimum wage to workers at Sea-Tac International Airport as well. At the time of the high court's ruling, proponents of the minimum wage said that 1,300 workers in SeaTac's off-airport transportation and hospitality businesses were earning the minimum wage of $15 per hour.

Diving Into Data

Having established the context in which SeaTac's minimum wage increase came into effect, let's next consider the actual month-to-month data that shows how employment levels in the Puget Sound city has changed over time.

Which for us is pretty cool because the city level employment data that we're using is something that we didn't know existed until Mark Perry linked to it shortly after he had discovered and used it to correlate large numbers of job losses in Seattle that took place in the months after that city's minimum wage hike took effect on 1 April 2015.

Since he had already looked at Seattle, we thought we'd look at another city in the Puget Sound area that made national headlines with its own minimum wage hike: SeaTac.

The data is based on detailed information that the U.S. Census Bureau collects each month as part of its Current Population Survey, where the Bureau of Labor Statistics then takes the employment data to craft its estimates of the nation's employment and unemployment levels. What we found that was new to us is that the BLS now provides access to its employment data down to a level that gives information for cities with 25,000 residents or more. Previously the closest we could get to city-level data was to consider what the Census calls a Metropolitan Statistical Area (MSA), which often includes other cities in its totals.

Because the raw data is not adjusted for seasonality, we accounted for that factor by calculating the trailing twelve month average of the reported employment data going all the way back to 2005. Because we wanted to assess the impact of SeaTac's minimum wage on its employment level, we also needed to establish a counterfactual, something that would give us an idea of what SeaTac's employment would have been if not for whatever change came about because of its action to increase its minimum wage.

Here, we took the average trend in the employment level that existed in the period from June 2006 through January 2008 as our counterfactual indicating what typical job growth was like for SeaTac outside a period of recession and abnormal (for Washington state) minimum wage hikes.

On this latter point, we should note that Washington state's minimum wage has been considerably higher than the U.S. federal minimum wage for a long time, where it has been adjusted for inflation in each year since it was first elevated above the federal level in 1999. As such, Washington state has not been affected by federal minimum wage hikes, and until 1 January 2014, the applicable minimum wage in SeaTac was Washington state's minimum wage, which was $9.19 per hour in 2013 and would have increased to $9.32 per hour on 1 January 2014.

Having now described where we got the data, and how we handled it for our analysis, which should be enough for any competent analyst to follow and replicate, let's next look at the chart we generated showing SeaTac's evolution in employment levels from January 2005 through December 2015.

Having generated this chart (without any annotations), one of the first things we did was to compare it to the chart that Mark Perry had generated for Seattle, which is where we noticed a problem.

Problems with the Data

Visually comparing the most recent data in the two charts, we couldn't help but notice that the nonseasonally adjusted data in both charts appeared to be following the exact same pattern, tick for tick, in recent months.

Meanwhile, older data in the charts did not. Our next step was to take Seattle's employment data and use it to produce a chart similar to the one we created for SeaTac's employment levels.

We see in the period before January 2014 that the trajectory of Seattle's employment level over time is very different from SeaTac's employment. We also see that the period since January 2014 appears to follow a nearly identical pattern.

Our next step was to take the trailing year average data we produced for each city's data to calculate the annualized growth rate for employment from each month to the next. The following chart reveals what we found.

In this chart, we see that SeaTac and Seattle follow similar, but different trajectories in the period preceding January 2014, and identical trajectories in the period from January 2014 through December 2015.

Although we focused on the five year period from January 2011 through December 2015, the same pattern can be seen in older data - interspersed with what appear to be population adjustments for SeaTac's data at five year intervals between December and January, which affects the data for the years 2000, 2005 and 2010. Data for SeaTac only extends back to January 1999.

What that tells us is that for the periods where both SeaTac and Seattle's month over month growth rates for their employment levels are identical is that the data jocks at either the Census Bureau or at the Bureau of Labor Statistics have not disaggregated the city level data from the overall MSA-level data.

The bad news is that this means that city-level data for cities within the Seattle MSA region is only meaningful during periods where it has been ungrouped from the whole. It may be that this level of granularity is something that comes about after the data goes through revisions - much like how the Bureau of Economic Analysis updates its estimates of GDP even years later as it accumulates more detailed information.

That affects Mark Perry's analysis of the impact of Seattle's minimum wage hike since it went into effect on 1 April 2015, as the data he used would not appear to have as yet gone through an update that would allow it to be differentiated from the full Seattle MSA. Things like layoffs in other cities in the Puget Sound region, such as Everett, Redmond, Renton and others are mixed into the data.

That's not an error on Perry's part, because there is no warning at the BLS' site interface that would alert one to that situation existing in the data. We only found it because we just happened to notice the similarity in recent data between his Seattle chart and our SeaTac chart and dug into it.

The good news is that the data for SeaTac's employment in the period before 1 January 2014 has been disaggregated from the whole Seattle MSA, so we can use it to explore some of the impact of that city's minimum wage hike.

Minimum Wage Job Loss in SeaTac

Because it's probably scrolled off your screen by now, let's look again at our chart for SeaTac's employment level....

In the chart, we see that there is a clear break in the trend for employment in SeaTac, which coincides with the May 2013 announcement that SeaTac voters would be voting on the city's Prop 1 ballot measure increasing its minimum wage to $15 per hour on 1 January 2014.

Before May 2013, we see that employment growth in SeaTac was growing at the same rate it did during our selected non-recession counterfactual period (June 2006 through January 2008). By December 2013, the data suggests that SeaTac resumed growing at its counterfactual employment growth rate. In between, SeaTac appears to have lost approximately 100 jobs.

Meanwhile, in Seattle, we see that employment was steadily rising throughout this period at the same pace as Seattle's counterfactual employment growth rate. That confirms that some factor was very definitely affecting employment in SeaTac at the time, but not employment in Seattle.

Since SeaTac city officials were unaware of any business closures or layoffs in the city related to its minimum wage hike, it is highly likely that the reduction in jobs at the $15/hour minimum wage level were spread among many businesses and quite likely achieved through attrition, where the employee leaving a job was simply not replaced.

Moreover, it suggests that employers in SeaTac were very forward-looking in using the period before the minimum wage hike would take effect to thin out their payrolls so they wouldn't have to lay them off afterward.

And since no airport workers were impacted by the city's minimum wage hike until August 2015, the reduction of 100 workers out of SeaTac's remaining 8,400 workers represents a 1.2% decline in the city's employed population. In January 2014, 1.2% jobs lost nationally would be equivalent of some 1.74 million jobs (1.2% * 145,092,000) being lost in 7 month period, which would constitute a recession by any other name.

Something to keep in mind at this point is that what the city-level employment data is really measuring is not jobs in SeaTac, but rather, the employment of people who live in SeaTac, Washington. As we saw in our earlier discussion, we know that at least 400 of the 1,500 people who would be directly impacted by SeaTac's minimum wage hike to $15 per hour actually lived in SeaTac, so if any job losses among the directly impacted minimum wage earners in the city were truly randomly distributed, minimum wage earning SeaTac residents would have a 26% chance of being included among them. The math that applies then is simply the statistics where small sample sizes are involved.

Speaking of which, if you paid close attention to our narrative above describing SeaTac and its minimum wage hike, which we've summarized directly from contemporary news sources, you will already recognize that at least 200 jobs disappeared in the city, which is the difference between the counts of 1,500 and 1,300 for the number of people employed at that minimum wage in SeaTac between 2013 and 2015.

Shall we split the difference and call it 150 job losses in SeaTac that may reasonably be attributed to the city's 63% increase it its minimum wage, plus or minus 50?