I'm looking for US hiring data at national level from 1947 to 2018. I'm interested in data similar to what the BLS website has available for after 2000.
The Job Openings and Labor Employment Survey (JOLTS) was launched in the year 2000, so there isn't any older data from this source.
This survey replaced an older one called the Labor Turnover Survey, which lasted from 1954 to 1981. This data was published in a report from the BLS called Employment and Earnings on a monthly basis. But if you pick one from each decade the report has an annual time series that covers the previous ten years. They're archived by the St Louis Fed here:
Here's an example of the last report that contained that data, from 1982:
This isn't great as there are two decades missing and the methodologies are a bit different. I went through a copy of the stats abstract from the mid 1990s to see if there was something comparable for the years in-between for hires or turnover, but didn't see anything.
Another possibility would be to use the Current Employment Statistics, which is a much longer series that stretches back to the 1930s (which Barry mentioned and linked to). The JOLTS page has documentation that high lights the difference between the two datasets:
"The Job Openings and Labor Turnover Survey (JOLTS) measures hires and separations on a monthly basis. If separations are subtracted from hires, the difference represents an implied employment change. The Current Employment Statistics (CES) survey provides net employment change. The JOLTS implied employment change and the CES net employment change are conceptually similar and are expected to track well with each other over time."
So it's a compromise. Net change is not the same as hires, but if you go with the CES you can get a long time series. Or you use JOLTS with the definition you want and a shorter time series, with the option of digging out the LTS data from those reports to extend the series a few decades with a gap in between.