The country counts jobs, but not exits

Each quarter, the Central Bank publishes how many people are employed in the Dominican Republic. In the first quarter of 2026, the figure was 5,236,178 workers, a year-over-year increase of 118,631 net employed. The figure is solid, the source is impeccable, and the reading almost everyone makes of it is the same: employment is growing.

That reading is correct and, for a company that needs to understand its own turnover, it is almost useless.

The number is net. It measures the balance between everything that entered and everything that left the labor market over twelve months. A balance of 118,631 is compatible with almost any scenario of movement beneath it: there may be mass hiring alongside mass departures, or a market that is practically still. The official figure does not allow the two to be distinguished. It counts the inventory at the close. It does not count how many times that inventory moved to get there.

And the inventory moves more than the balance suggests. The Central Bank itself publishes it, in a section almost no one reads through this lens. The ENCFT labor transition panel follows the same people from one year to the next. Of every group of employed people in the first quarter of 2025, 6.5% had left the labor force twelve months later and an additional 1.8% had become unemployed. Close to one in twelve employed workers stopped being employed within a year. That is movement, measured in a primary source, and it does not appear in the job-creation headline.

For a company, the difference between those two ways of reading the market is the difference between knowing there is traffic and knowing how fast its own car is going.

It is worth being precise, because the easy claim would be that exits are not recorded in the Dominican Republic. They are recorded. The Ministry of Labor, through the Integrated System of Labor Records, receives the personnel movements of formal companies —hirings and terminations— through its regulatory forms. The gross entry and exit data exists within the State's system.

What does not exist is the reading. That flow is not published as an indicator a company could use to benchmark itself, nor broken down by sector, nor converted into a rate. The information is captured and remains illegible to whoever would have to make decisions with it.

When a Dominican company looks for a turnover benchmark to know whether its own is high or low, it finds private estimates from consulting firms, built on their own samples and methodologies that are not always published. And it finds that they do not agree with one another.

One general market measurement has placed average turnover at around 8%, with a range running from 4% to 15% depending on the sector. A more recent regional report places the Dominican Republic in a considerably higher band. The two references differ by a factor close to two.

The discrepancy does not mean one is wrong. It means there is no shared standard for what counts as an exit —whether voluntary separation is included, whether the end of a temporary contract is counted, whether the probationary period is in or out— and for how to calculate the rate. Without a common definition, two honest measurements of the same country arrive at numbers that contradict each other. That is the real state of turnover information in the Dominican market: perception exists, isolated estimates exist, and the standard that would make them comparable is missing.

The direct cost of a separation is budgeted and absorbed. It is visible: it appears on the payroll, it has a line item. The rest of the cost is paid without being recorded. The knowledge the departing person did not manage to transfer. The productivity deferred while the replacement learns. The load that falls on the teams that remain and sustain the operation during the transition. None of those three has a line item, and that is why it rarely appears broken out in an internal analysis.

What is not measured precisely is not corrected precisely. A company that manages its turnover for years without the data to understand it ends up attacking symptoms: it raises an incentive, replaces a supervisor, redesigns onboarding. Something improves for a few weeks, and the pattern returns, because the pattern was never understood. The decision is made on an impression, and an impression does not distinguish cause from noise.

The problem has a technical origin before a matter of will. Companies capture the entry with discipline —contract, onboarding, payroll— and leave the exit in a supervisor's memory. The country measures the inventory with statistical rigor and leaves the flow unread. The information that would allow the cycle to be interrupted is not lost: it is unstructured, scattered across records no one has connected to turn them into something an administration can use.

Building that structure is work, and it is the work Nexurium does. Before solving a company's turnover, one has to be able to measure it. Before measuring it, someone has to build the infrastructure that captures it. That is the order, and it is where everything else begins.