n.V1-5.08| THE CAPACITY THAT LEAVES

01| The familiar complaint about talent in this sector is dispersion: capable people scattered across a million small organisations and causes, each doing a little, none reaching the scale where anything changes.

02| That version is intuitive, widely repeated, and unmeasured. No study we can find demonstrates that talent dispersion across many small organisations reduces aggregate impact. It may be true. It is not evidence, and it should be stated as a hypothesis or not stated.

03| What is measured is a different movement, and it is worse. Within a country, the aid system does not mainly spread capable people thinly. It concentrates them — pulling them out of state institutions into project delivery, and pricing them by nationality rather than capability.

04| The clearest measurement comes from health, where professionals are countable and their movements traceable. Research on Mozambique documented what happened as external assistance surged through global health initiatives using NGOs as their primary delivery channel. Roughly a quarter of the country's physicians — 25.5 percent — had left the public sector. Of those who left but stayed in the country, 62.4 percent were working for NGOs, 21.2 percent for donor agencies, and 12.4 percent in the private sector.

05| The finding that reframes the whole discussion of brain drain is this: internal migration accounted for more physician loss to the public system than emigration did. More doctors were lost to organisations operating inside Mozambique than to hospitals abroad.

06| The drivers were not mysterious — higher salaries, better conditions, more rewarding career development. And when public sector health workers were asked about vertical funding arrangements, the exodus of staff into international and private organisations was the single greatest problem they named: widespread, persistent, and causing lasting damage to the national system.

07| That connects directly to fragmentation, and they are the same phenomenon seen from two angles. Knack and Rahman's mechanism predicts exactly this: as the donor field splinters, each donor's incentive to extract capability from the state strengthens and its incentive to build any strengthens not at all.

08| The second measured pattern concerns how the sector values people once it has them. Most international organisations run a dual salary system. National staff salaries are set against the national median wage; international staff salaries against a global scale or the labour market of the organisation's home country. Two people doing comparable work with comparable qualifications are priced against two different economies.

09| The reported gaps are large. Studies put the ratio anywhere from four to fifteen times, with national staff commonly paid around four times less on average than international counterparts of similar qualification and experience, and disparities as high as ninefold reported in some Pacific contexts. One recent analysis of Ukraine found international staff costing roughly five times national staff — and seventeen times the cost of staff at local NGOs.

10| The documented consequences are what anyone would predict: dissatisfaction, reduced productivity, high turnover. And a second-order effect that closes a loop. Talented national staff, facing a ceiling set by their nationality rather than their ability, leave the country for better-paid work elsewhere. The dual salary system, designed to make international deployment affordable, feeds the external brain drain the sector separately laments.

11| So the movement runs in two stages. Capable people are drawn out of national institutions into internationally funded delivery, and then a portion of them are pushed out of the country altogether by a pay structure that will not value them at home.

12| The case against reading this as villainy is serious and should be made at full strength. An organisation that offers a doctor better pay and better conditions is not wronging that doctor. The doctor is better off and freely chose. The patients treated through the vertically funded programme are real patients, and the health outcomes those programmes produced — in HIV treatment above all — were real and large. A public health system that cannot retain its staff has problems that predate the arrival of any NGO, including salaries its treasury cannot or will not raise. And an organisation that cannot recruit internationally at internationally competitive rates cannot operate in difficult places at all.

13| Here is the gap, and it is why this remains an argument rather than a settled finding. Nobody has netted it out. The health gains delivered by vertically funded programmes are documented. The capacity losses those programmes caused in national systems are documented. No study we can find puts them on one ledger and says which is larger, over what horizon, in any specific country. The two literatures barely reference each other.

14| That absence is not a conspiracy. It follows from everything else in this book part: the gains are attributable to a funder and therefore get counted, while the losses are diffuse, fall on an institution nobody is reporting on, and appear years later.

15| So the supportable claim is narrower than the intuition and considerably sharper. Aid does not scatter talent. It moves it, systematically, from the institutions that have to last into the projects that do not.