How to read what's really happening in a weird US jobs report

New York(CNN Business) When a surprisingly bad jobs report comes in like the one released on Friday morning, it's tempting to declare that the economy now looks radically different from how it looked just before the report.

The latest report, which some economists called "bizarre," appeared to show a sudden slowdown in hiring, with the economy adding just 20,000 jobs in February — a far cry from the 311,000 jobs added in the prior month and the weakest job growth since September 2017.

At first glance, it looked like the economy took a sudden turn for the worse.

But there's a lot more to it than that.

The Labor Department's Employment Situation report contains a trove of data points that, when analyzed as a whole, can turn what looks like a bad report into a not-so-terrible one.

1. No single jobs report matters that much.

If something looks like a blip, it's probably a blip.

We need to look at at least three months of data to conclude that a trend is real. Remember, all of these numbers will be revised over the following two months, potentially by a lot.

That's why anyone heralding a radical shift because of any one report is probably wrong.

In this case, the February report showed monthly job gains averaged 186,000 over the last three months, marking a modest pace of hiring.

2. Other data besides the jobs report are also important.

This is especially true on wages.

The February jobs report showed average earnings grew by 3.4% over the last year — a strong number that could indicate workers have the upper hand in a strong job market.

But average hourly earnings are just one measure of how much workers are making. Separately, the government publishes another report on compensation, which tracks how much companies pay for their workers, including benefits.

It's also a good idea to factor in inflation to understand what paychecks actually have the power to buy. BLS publishes a monthly report on real earnings. As of January, it showed hourly earnings were up 1.7% over last year, once inflation was included in the mix.

3. The jobs report is really two reports.

The Labor Department conducts two surveys to compile the jobs report: One of employers, from which it draws the total number of jobs gained or lost. The other is of households, from which it draws the unemployment rate.

In February, the household survey showed the unemployment rate fell to 3.8% as fewer workers were counted as unemployed. That's in part because federal workers returned to work after being furloughed during a partial government shutdown in January.

The BLS has a handy explainer on the differences between its surveys of employers and households. One of the most important is that the household survey is much smaller, with a sample size of only about 60,000, so the data can be noisier.

4. A higher unemployment rate can actually be a good thing.

The unemployment rate measures the share of people who are out of work and want a job. So if more people tell the Labor Department that they've looked for a job in the past four weeks, the rate can go up.

If that happens, the labor force participation rate could also go up. That measures the share of people who are working or want to work.

Sometimes people say they're looking for a job because wages are rising and it seems like there are better opportunities out there. They'll be counted as unemployed, but their participation is a good sign for the economy.

5. Work isn't binary.

There are many shades of having a job.

For example, the household survey asks people whether they were employed part time because they wanted to be or because they couldn't find a full-time job. Then there are the "marginally attached" workers who looked for work in the past year but not the past four weeks. And the Labor Department measures "discouraged workers," who haven't looked for work because they don't think any is available.

All of that goes into something called the U-6 rate, which gives a more comprehensive look at "labor market slack," or how many workers remain ready to do jobs if the right ones present themselves.

As of February, that rate was 7.3%, its lowest level in about 18 years.

6. Sometimes seasons matter.

Labor Department data comes in two types: Seasonally adjusted and not seasonally adjusted. Seasonal adjustment just means that the statisticians have tried to filter out anything that might be a repeating pattern having to do with the time of year, including major holidays, weather or school schedules.

If a data series is not seasonally adjusted, don't look at the month-to-month change, only look at the change from a year ago. This is also true of the state and local employment and wage numbers the BLS puts out.

With an increasing number of crazy weather events, BLS isn't able to totally filter out the usual impacts of snowstorms and hurricanes. That's why they also ask workers if they couldn't get to work because of the weather, which sometimes has a big impact on the overall payroll number.

This February, 390,000 people who usually have jobs, said they didn't work because of weather. Another 1.9 million people who usually work full time, said they worked part time instead, because of weather.

7. Industry categories don't always mean what you think they mean.

The Employment Situation breaks out job gains into broad industry "supersectors." It's best not to make assumptions about what these mean. "Business and professional services," for example, has been one of the largest job gainers in recent years.

That's people like lawyers and accountants, but also staffing companies and waste haulers. If you want to understand what professions are really growing or shrinking, you need to look under the hood.

In February, department stores cut 9,000 jobs and heavy engineering firms lost 13,200 jobs. The health care industry added 20,800 jobs.

8. FRED is your friend.

The BLS website's data tools are useful, but they can be clunky and hard to navigate. To get a quick picture of longer-term trends, fire up the St. Louis Federal Reserve's FRED website, which is an incredibly easy-to-use data portal that offers lots of tutorials on how to marshal and manipulate almost every bit of information the federal government produces (and more).