Thailand is usually held up as the country that cracked rural doctor shortages. Since 1994, its Collaborative Project to Increase Production of Rural Doctors has recruited medical students from rural backgrounds, trained them partly in provincial hospitals and bound them to three years of public service. Breaking the contract meant a fine the study put at the equivalent of about US$12,500.
It is one of the most serious rural workforce programmes anywhere, and it works. A cohort study published in Human Resources for Health in 2015 followed more than 7,000 Thai doctors who graduated between 2000 and 2007. The rural-track doctors were more likely to stay in rural hospitals than their peers. Their predicted median time in a rural hospital was 4.2 years, against 3.4 years for doctors from the normal track.
Read that again. The flagship programme, with rural selection and rural training, bought about ten extra months. And the compulsory service was not what made the difference: according to the study, the same three-year obligation and the same fine applied to normal-track graduates too. Compulsion kept both groups in place for the bonded period. What the rural track added came after it. By the end of follow-up, 29% of the rural-track doctors were still in rural hospitals, against 18% of the others.
That is a genuine improvement, and Thailand deserves credit for it. It is also a warning. If the best-designed programmes add months rather than decades, then the goal of “retaining” rural health workers, as usually framed, may be the wrong goal.
The scale of what is being asked
The WHO’s 2021 guideline on rural and remote health workforce notes that almost half the world’s population lives in rural areas, and that in some countries rural populations have access to ten times fewer health workers than urban ones. Behind that sits a global shortfall that WHO’s most recent projections put at around 11 million health workers by 2030, concentrated in low- and middle-income countries.
Rural facilities are also competing with an international market. WHO’s 2023 health workforce support and safeguards list named 55 countries where health worker density and service coverage are so low that active international recruitment from them should be avoided. Thirty-seven of those are in the African region. A nurse in a rural district is not choosing only between her village and the capital. She may be choosing between her village and a hospital in another continent.
In that context, expecting individual health workers to stay rural for a whole career is not a policy. It is a hope.
What ten months is worth
It is tempting to dismiss the Thai gain as small. That would be a mistake, and working through the arithmetic shows why.
Suppose a province needs 100 doctors in its rural hospitals at all times. If each doctor stays an average of 3.4 years, the province needs to recruit roughly 29 new doctors every year just to stand still. If the average rises to 4.2 years, the requirement falls to about 24. That is five fewer recruits a year, every year, from the same pipeline, and five fewer periods in which a post sits vacant between one doctor leaving and the next arriving. Over a decade, the difference adds up to a substantial number of doctor-years of rural care.
Seen this way, the Thai programme is a success on exactly the terms this post argues for. It does not keep doctors forever. It makes the flow cheaper to sustain. The mistake is not in the programme but in the way such programmes are sold, as if the aim were a lifetime of rural service and anything less were failure.
The same arithmetic also shows the cost of vacancies. If it takes six months to fill a rural post after a doctor leaves, and the average stay is under four years, the post is empty for something like an eighth of the time. Faster replacement can be worth as much as longer retention, and it is usually cheaper to buy. Streamlining recruitment paperwork, pre-assigning incoming graduates to posts that will fall vacant, and overlapping outgoing and incoming doctors for a few weeks are administrative fixes, not new programmes. They rarely attract attention because nobody launches them with a ribbon.
Why retention thinking goes wrong
Most retention strategies treat leaving as a failure: of the worker, the incentive package or the programme. So the response is to add more incentives, extend bonding periods or raise penalties.
There is a limit to how far that goes. Compulsion can fill posts, but it tends to produce resentful staff who count the days to release. Financial incentives help, but rarely compete with the combined pull of urban schools for children, spousal employment, specialist training and professional peers. The factors that drive people out of rural practice are often about family and career, not salary.
And there is a quieter problem. A retention target measured in years of individual service rewards keeping people in place, even when they have become isolated, deskilled and burnt out. Nobody benefits from a doctor who stays for eight years and stops learning after two.
What compulsion does, and does not do
The Thai data carries a quieter lesson about compulsory service. Bonding worked in the narrow sense: it kept doctors from both tracks in public service for the bonded period. What it could not do was make anyone stay once the bond expired. The extra retention came from who was selected and where they were trained, not from the contract.
Compulsion also has side effects that are easy to miss. A fixed fine is a heavier burden for a student from a poor rural family than for one whose parents can pay it, so bonds can end up binding exactly the people the programme most wants to recruit while others simply buy their way out. In some countries, the prospect of compulsory rural posting may push students toward private medical schools or abroad. And a doctor serving a bond resentfully may do the minimum, which is its own kind of vacancy.
None of this means bonding should be abolished. A period of guaranteed service can stabilise staffing while other measures take effect, and some doctors discover during it that they like rural practice. But it is best understood as a bridge, not a retention strategy. What happens when the bond ends is decided by everything else: the team, the training path, the family’s circumstances and whether rural work feels like part of a career.
The counterargument: continuity matters
The strongest objection to the flow model is continuity of care. A rural doctor who stays for fifteen years knows the families, the history and the local pattern of disease in a way that a series of four-year postings never can. Continuity is associated with better care in primary care generally, and a model that assumes turnover could quietly give up on it.
That objection is fair, and it is why the flow model should not be read as indifference to long service. Some people do stay for decades, and they are extraordinarily valuable. Systems should make it easy and attractive for them to stay, with career progression that does not require leaving, and with recognition of the experience they carry.
But continuity does not have to rest on the doctor alone. In a well-built rural team, the nurses, clinical officers and community health workers who are recruited locally often stay far longer than the doctors, and they carry much of the knowledge of the community. Good records, stable local staff and structured handovers between doctors preserve more continuity than hoping each new doctor will stay forever. The long-staying doctor is a bonus to plan for, not a foundation to plan on.
Design for flow, not for stock
A more honest approach accepts that most rural health workers will serve for a period and then move. The design question becomes how to make that period productive and how to replace people smoothly.
That shifts attention to things retention programmes often neglect:
- Pipeline volume. If each person serves four years on average, the system needs a steady annual intake of new rural workers, sized to that turnover. Thailand’s model of selecting students from rural areas matters here more for filling the pipeline than for keeping any one doctor.
- Rotation with a return path. Structured rotations between rural and urban facilities, with guaranteed access to specialist training afterwards, can make rural service a career step rather than a dead end.
- Team resilience. A rural health centre that depends on one doctor collapses when that doctor leaves. One built around nurses, clinical officers and community health workers, with the doctor as one member of a team, absorbs turnover far better.
- Local cadres. Health workers recruited and trained close to home, including mid-level providers, tend to have fewer reasons to leave and fewer opportunities to emigrate.
None of this abandons retention. The WHO guideline still recommends a bundle of measures across education, regulation, financial incentives and personal support. But it changes the success measure from “how long did each person stay?” to “was the post filled by a competent person, continuously?”
What this asks of ministries, funders and training institutions
For ministries of health, the practical shift is in planning. Workforce plans built on turnover assumptions, with training intake sized to replacement need, are more honest than plans that assume each post, once filled, stays filled. Ministries can also track time-to-fill for vacant posts as closely as they track retention, because it is often the more fixable number.
For medical and nursing schools, the lesson from Thailand is that selection and training location matter. Recruiting students from rural areas and training them partly in rural hospitals appears to add real months. Those are choices schools can make without waiting for new money.
For donors and development partners, the temptation is to fund incentive schemes for individual workers, which are easy to announce and hard to sustain after the grant ends. Supporting the training pipeline and the team structure is less visible and more durable. And for destination countries recruiting internationally, the WHO safeguards list is a reminder that each nurse hired from a listed country leaves a gap that someone else’s rural district has to fill.
The trade-off governments dislike
The flow model is more expensive up front. Training more people than you need at any one moment feels wasteful to a finance ministry. Building teams instead of posting single doctors costs more in salaries. Rotation schemes need administration.
The alternative is cheaper on paper and costlier in practice: vacant posts, locum spending, closed clinics and a rural population that learns not to rely on the public system. Those costs are real. They simply sit in a different budget line, or in the health of people who are not counted.
The takeaway
Rural retention programmes deserve continued investment, and the best of them, like Thailand’s, show what serious effort looks like. But the evidence says they extend service by months, not careers. Health systems that plan around that reality, with bigger pipelines, stronger teams and rotation that leads somewhere, will keep rural clinics staffed. Systems that keep promising lifelong rural doctors will keep being disappointed.
In your experience, what has actually kept health workers in rural posts longer: money, training, family support or something else? Tell us in the comments.
