Sample

A sample of 2M white-collar workers, split evenly between remote workers and in-office workers.

Methodology for determining remote vs in-office

The best available proxy for determining remote versus in-office workers is to check if an individual lives in the same metro area as their employer's office, including company HQ and other office locations. Using a list of office locations for the companies in the sample, we then compared the location data for the individual to location data for company office location(s). If the distance was >50 miles or fell outside of the metro area (see: special cases such as Raleigh-Durham-Chapel Hill which has a radius greater than 50 miles but would still be a realistic commute) the person was deemed “remote". If the person was within the 50-mile radius or metro area, they were assumed to be spending a meaningful amount of time in the office. While there are edge cases, this is a generally accepted proxy for determining an individual’s likelihood to spend time in the office versus fully remote.

Methodology for determining layoffs

From a sample of 2M white-collar workers (split evenly between remote workers and in-office workers), we observed 599K+ (29.95% of the sample) job change events in 2023 (company change or internal change). Of these changes, 373K+ (18.6% of the sample) were company changes. For the people in the sample who had a company change event, we then looked to see if the individual spent more than 60 days without starting a new job. If the individual left a company and did not start a new job within 60 days, we classified this job change as involuntary or a “layoff”. From our sample of 2M people, 168K+ (8.4% of the sample) people were laid off.

Notes on layoff rates

The data presented above are annual rates for white-collar workers in 2023. Our data shows a lower annual layoff rate (~8%) versus the annualized BLS number for 2023 (rough math puts it at ~12% = 1% x 12 months). This difference is likely due to three factors:

1) Given Live Data’s classification of involuntary job changes (white-collar professionals who leave a company and do not start another white-collar role within 60 days), we haven’t been able to classify people from late November and December yet since they haven’t yet met the 60-day criteria. Modeling the data for late November and December, the Live Data layoff rate for 2023 gets even closer to the annualized BLS number.
2) Live Data’s focus is white-collar workers. While we certainly see a large amount of turnover within one year for the white-collar workforce, it is reasonable to expect that the increased time and resources it takes to recruit, train, and onboard a white-collar worker (vs blue-collar or service industry workers) that there is there are fewer layoffs and company changes for this cohort than for the US workforce as a whole.
3) Our best proxy for determining layoffs requires someone to stay laid off for 60 days. While the majority of white-collar professionals who are laid off take at least this much time to find their next professional role, some people are hired within 60 days.

Methodology for determining quits

From a sample of 2M white-collar workers (split evenly between remote workers and in-person/hybrid workers), we observed 599K+ (29.95% of the sample) job change events in 2023 (company change or internal change). Of these changes, 373K+ (18.6% of the sample) were company changes. For the people in the sample who had a company change event, we then looked to see if the individual spent more than 60 days without starting a new job. If the individual left a company and started a new job within 60 days, we classified this job change as a voluntary job change or a “quit”. From our sample of 2M people, 205K+ (10.3% of the sample) people quit in 2023.

Methodology for determining promotions

From a sample of 2M individuals (split evenly between remote workers and in-office workers), we observed 599K+ (29.95% of the sample) job change events in 2023 (company change or internal change). Of these changes, 225K+ were title changes at the same company. We then looked to see if their new job level was higher than their previous job level using an ML solution for title-to-level classification. If the individual’s new title was a higher level than their previous title, we classified this job change as a promotion. From our sample of 2M people, 94K+ people were promoted.

Notes on the source data from Live Data Technologies

Live Data Technologies's standard process (our patented IP) is generally referred to as SERP analysis. We query the major search engines (Baidu, Yandex, Bing, Google, etc.) for information on people and their employment. A shorthand way of thinking of this is that we are prompt engineering the search engines.

All of our data is sourced through these sources and is publicly available. We assess ALL the info that comes back from each query to the search engines. We use a proprietary process to monitor the current company and title for 88M+ people on at least a twice-monthly basis, and as such pick up a lot of job changes monthly. This means we have a) the most recent employment data for the white-collar workforce and b) a continuous stream of job change events. This allows us to report on movement at the person, title, function, job level, company, and industry levels.

‍