Conservative Twitter is blaming the “deep state” for the lackluster US jobs report ?

The US economy added 138,000 jobs in May, down from 174,000 in April and below analyst predictions of 185,000 jobs, the US Bureau of Labor Statistics (BLS) said today.

These stats are released monthly by the nonpartisan BLS—or as conservative Twitter saw it this morning, surreptitious agents of the so-called “deep state” hoping to undermine Donald Trump’s presidency.

The conspiracy-stoking tweets, sent just after the lackluster jobs report was released, claimed the BLS statistics were wrong and that political holdovers from the Obama administration were somehow “spiking the numbers” to make Trump look bad. The 2016 presidential campaign also saw routine economic statistics called into question as potentially cooked for partisan gain.

Today, conservative online talk radio host Bill Mitchell and others pointed to discrepancies between the BLS numbers and the survey released by US payroll management company ADP, which said 253,000 private-sector jobs were added in May.

Okay, we’ll bite.

Why is there a difference between the ADP and BLS jobs numbers?

ADP releases its report two days before the official BLS numbers come out each month. Its estimates of job growth are based on data from its clients (about one-fifth of U.S. private payroll employment), past BLS data, unemployment insurance claims, surveys of consumer confidence, oil prices, and more (pdf).

The BLS surveys businesses and households monthly to generate the official estimates of jobs gained and lost. These results are revised periodically to reflect more complete information and a better understanding of seasonal factors, among other things.

Back in 2012, Wall Street Journal (paywall) columnist Steven Russolillo cautioned readers that there is often a “wide divergence” between ADP data and government numbers. “It’s been well documented how ADP’s report, which measures private-sector job growth, has an uneven track record of predicting the nonfarm payrolls data,” he wrote. “Some months it’s spot on, others it’s wildly off base.”

Is it unusual to see a significant disparity between the ADP and BLS reports?

No.

The difference between the number of new private-sector, non-farm jobs reported by ADP and the BLS every month is greater than 40,000 about half the time, according to a CNBC analysis of the figures since November 2012.

“The ADP may be a valuable measurement reflecting economic fundamentals, but for an investor looking to predict what the market will do… the headline number is probably not very useful for specific predictions — such as whether the BLS will beat consensus estimates,” CNBC wrote.

Why do BLS figures get revised each month?

Today, the BLS released May job numbers for the first time, and also said there were 50,000 new jobs in March and 174,000 in April, down from estimates of 79,000 and 211,000, respectively, in last month’s report. These revisions indicate that the labor market isn’t as strong as it may have seemed.

As is common practice with economic data, the BLS periodically revises its numbers—twice in the months after an initial data release, and then again in a more comprehensive annual benchmarking process.

Read next: 10 charts for tracking whether Trump is delivering on his economic promises