35 is 15% larger than 30. Our measurements for unemployment are precise enough that it’s not just margin of error.
At the moment, that’s a change of 16 million people. A century ago, it was 5 million people. Both of those numbers are large enough that they absolutely would get attention and public discussion.
If something trends give or take 30% for 80 years or more, a +/-3% variance is a fucking rounding error. I don't care if it's home ownership or unemployment or whatever the number is, it is a rounding error not worthy of concern.
It's like the lowering and rising of the tides. It goes up and down, but the sea level remains basically the same overall; thus, nobody cares about the exact tide at an exact time for figuring out sea level because it's a rounding error.
Put more bluntly: The trend has not changed, what is there worthy of my concern?
Because the economy is a complex system which is constantly changing, and it’s important not to lose track of why we measure it. A 5% charge likely isn’t evenly distributed, and the real impacts are often masked – if a major industry moves offshore, there are inevitably regional impacts more significant than the national average and if a bunch of people get new jobs working at McDonald’s or delivering Amazon packages it matters that they and their communities are poorer even if it looks like the unemployment rate is back to where it was 3 years ago. The minor fluctuations are the cue to look at the responsible factors, not to say “meh, it’s 15%, not 50%” and ignore it.
At the moment, that’s a change of 16 million people. A century ago, it was 5 million people. Both of those numbers are large enough that they absolutely would get attention and public discussion.