We all know that math scores have been scandalously trending downward for many years, but the folks in the government should at least be able to count.

        We're finding more and more evidence that the statistics the government is releasing to the public are increasingly suspect and unreliable. It seems like the errors are not random but perhaps manipulated for political advantage. Judge for yourself.

        Let's start with crime statistics. Former President Donald Trump said in the debate that crime is out of control, and Vice President Kamala Harris countered by citing government statistics from the FBI indicating that crime rates are falling.

        But Jeffrey Anderson, former director of the Bureau of Justice Statistics, finds a surge in urban violent crime since 2019. He writes in the Wall Street Journal that "the violent crime rate in 2023 was 19% higher than in 2019." The urban violent crime rate was up 40%, and urban property crime rate rose 26%.

        How can the Left keep saying crime is down? A big reason is the FBI figures are only measuring "crimes reported to the police." More than half of violent crimes are not reported, thanks to what Anderson calls a new era of "lax law enforcement policies" in urban areas. Police in big cities also have an incentive to undercount crimes to make their performance look better.

        Next, we have jobs data. The Bureau of Labor Statistics admitted last month that it has overstated job growth by more than 800,000 positions. And in just the last year the government has also overstated job growth by almost 500,000 from the original monthly headline numbers. This is an overcount of over 1 million. In 10 of the last 13 months, the errors were in the direction of announcing too many jobs.

        So President Joe Biden gets the gangbuster headlines, and the whoopsie daisy comes later when no one is paying attention.

        Those aren't just random errors. Was the Biden Labor Department finagling the data? Maybe.

        Then there was the decennial Census Bureau population count. The numbers from the 2020 census were wildly wrong, as the bureau admits.

        In an analysis issued in 2021 called the "Post-Enumeration Survey Estimation Report," the Census Bureau reported which states recorded overcounts of their population, and which saw undercounts. Florida, Texas, Tennessee and other red states were undercounted by some 1.5 million residents. The overcounting was in mostly blue states like New York and Minnesota. Again, was this just an accident?

        The miscount may have cost Republicans three electoral seats. This means the presidential election and control of the House of Representatives may be decided because of an error in counting heads.

        These government agencies are supposed to be politically independent, and historically, they have been filled with professionals devoid of bias. But when we see the errors all bending the data in the direction of benefiting one party, one has to wonder if this is deliberate misrepresentation.

        I hope I'm wrong and that these are innocent errors. But we live in an era where everything in Washington is hyper-politicized. Elections have become a blood sport. The saying is that "all is fair in love and war." And now add politics to that.

        Stephen Moore is a visiting fellow at the Heritage Foundation. He is also an economic advisor to the Trump campaign. His new book, coauthored with Arthur Laffer, is "The Trump Economic Miracle."