Summary
- RIPE NCC training, certification support and community fellowships serve valid capacity-building purposes, but scarce or selectively allocated benefits can create gratitude, access and visibility that matter in low-turnout association elections.
- Patronage risk does not require an explicit exchange of benefits for votes. It arises when candidates or officeholders can influence selection, timing, recognition or exceptions while beneficiaries cannot distinguish institutional entitlement from personal favour.
- The strongest safeguards are published eligibility rules, independent selection, documented recusals, aggregate distribution reporting, appeal channels and a campaign-period firewall between candidates and benefit decisions.
- Electoral legitimacy improves when members can audit the allocation system without exposing applicants' personal circumstances or treating every recipient as politically compromised.
The gift that is not supposed to be a gift
A membership association is expected to deliver benefits. That is not a corruption problem; it is part of the bargain. Members pay fees, accept contractual duties and support a common institution because collective provision can do what isolated operators cannot. The RIPE NCC offers training, technical learning, certification support, tools, meetings and opportunities to take part in a regional community. These activities can improve routing practice, strengthen operational security and help engineers understand the rules that govern Internet number resources.
An association that collected fees while returning no common value would have a different legitimacy problem.
The difficulty begins when an institutional benefit feels like a personal favour. A seat on an oversubscribed course, a funded journey to a meeting, a fellowship, an exam voucher, a speaking opportunity or access to a prominent mentor may be awarded under a general programme. Yet the recipient experiences the benefit through identifiable people: staff who selected the application, board members who championed the budget, community figures who offered introductions, or candidates who later remind voters what they helped create. The formal payer is the association. The emotional creditor may appear to be an officeholder.
That gap between institutional funding and personal attribution is the entry point for patronage. Patronage need not look like an envelope handed over in return for a ballot. In professional communities it is more likely to operate through recognition, travel, education and proximity. A person who was previously outside the central circle receives a funded opportunity, becomes visible, and develops understandable loyalty to those associated with the opening. No one needs to say how that person should vote. The obligation can remain implicit.
The issue is especially sensitive in an association election because the electorate is not the public at large. Voting rights follow membership and registration rules. Turnout is a fraction of the total service region, and active community networks can matter more than raw institutional size. A modest number of beneficiaries who become engaged voters, nominators, campaign volunteers or trusted messengers can have influence beyond their count. Capacity building and electoral mobilisation can therefore occupy the same social space even when their stated purposes differ.
The correct response is not to abolish benefits. Removing training or fellowships would punish newcomers and peripheral operators while leaving established networks untouched. It would turn an integrity concern into an incumbency advantage. The better question is whether the distribution of support is governed as a public institutional function: clear purpose, stable criteria, independent judgment, recorded conflicts, reviewable outcomes and no candidate control over individual awards.
Three different benefit systems
The phrase "membership benefits" can conceal important distinctions. The first category is a broad entitlement connected to membership. RIPE NCC materials state that members receive access to in-person training without an additional course fee, while online learning and webinars are widely available. Such an entitlement is comparatively resistant to patronage when every eligible member can use it under the same published conditions. The decision is largely automatic: membership establishes access, and scheduling determines when it can be exercised.
The second category is capacity-constrained support. A classroom has limited seats. A trainer can visit only so many locations. Exam vouchers, if offered in a fixed annual quantity, have a finite value. Meeting support has a budget. Even when the general programme is universal in principle, scarcity reintroduces choice. Someone decides which city receives a course, which language is prioritised, which waiting list moves first and whether an exception is justified. These are operational decisions, but repeated choices can shape who sees the institution as attentive and who sees it as remote.
The third category is selective sponsorship, including fellowships designed to bring new voices into the RIPE community. The RIPE Fellowship programme publishes eligibility conditions and describes a learning journey before, during and after a meeting. Selection is integral to the design because not every eligible applicant can receive funded participation. The benefit is also more socially powerful than a routine service. It includes travel support, coaching, introductions and a recognised identity as a fellow. Those features are valuable precisely because they accelerate belonging.
These systems should not be judged by one standard. A universal online course needs reliable access and fair language coverage. A limited classroom needs allocation rules and geographic reporting. A fellowship needs robust selection independence, conflict controls and privacy protections. Calling all three simply "benefits" encourages either complacency or overreaction. The governance risk lies in the discretion attached to each form of support.
There is also a legal and institutional distinction between the RIPE NCC membership and the wider RIPE community. The association has formal members and General Meetings; the community is open to people who may not represent a member. A fellowship aimed at community participation is not necessarily a dividend paid to voting members. That distinction matters because the route from benefit to ballot can be indirect. A fellow may later join a member organisation, become its registered voter, influence an authorised representative or acquire standing as a community endorser. Electoral effects do not depend on immediate voting eligibility.
The analytical task is therefore to follow influence rather than labels. Who sets the programme? Who chooses recipients? Who communicates the decision? Who receives public credit? Can a candidate intervene? Can rejected applicants challenge inconsistency? Are aggregate outcomes visible? Those questions reveal whether support operates as neutral capacity building or as a reserve of discretionary favour.
Patronage without a bargain
Classic vote buying requires an exchange: a benefit is offered on condition that a person votes in a particular way. Association patronage is usually harder to see because there may be no condition, no proof of the vote and no explicit instruction. Electronic ballots are secret, and a candidate cannot ordinarily verify compliance. It would be a mistake, however, to conclude that secrecy eliminates influence.
People reciprocate attention. A recipient may support an officeholder out of gratitude, trust or a belief that continuity will preserve the programme. An organisation that hosted a training course may regard the board as responsive to its region. A community organiser who helped identify fellowship candidates may gain standing that later translates into endorsements. These are ordinary social effects. They become governance concerns when access depends on discretion that is concentrated around electoral actors.
The absence of a spoken bargain can actually make the arrangement more durable. An explicit demand would offend community norms and create evidence. An implied relationship allows everyone to maintain an account of principled conduct. The officeholder says the programme served capacity. The recipient says the preferred candidate supports inclusion. Both statements may be sincere. Patronage works through the overlap.
The relevant test is structural, not psychological. It asks whether a reasonable beneficiary could believe that future access depends on maintaining favour with a candidate or board faction. It asks whether a reasonable unsuccessful applicant could suspect that political distance mattered. It asks whether the same criteria would have produced the same result if the application had come from a critic. The institution need not prove what anyone felt; it must show that influence was not available to be traded.
This is why published criteria alone are limited public evidence. A programme can display clear eligibility rules while leaving ranking highly subjective. "Potential to contribute" and "community engagement" are legitimate considerations, but they can also reward familiarity. A selector may prefer a known volunteer, a fluent conference speaker or someone recommended by a trusted insider. The result may reproduce an electoral network without any written political criterion.
Nor is equal treatment the same as identical treatment. A programme designed to reach underrepresented regions may properly direct more support to people who face higher travel costs, visa barriers or weaker employer funding. A patronage analysis should not condemn targeted inclusion merely because benefits are uneven. The question is whether the targeting follows an adopted purpose and objective evidence, rather than the preferences of candidates who can later harvest loyalty.
Why training is politically valuable
Training appears less political than travel sponsorship because its subject is technical. A routing-security course, IPv6 workshop or registry tutorial does not look like a campaign event. Yet technical education produces several assets that matter in association politics: competence, status, relationships and confidence to speak.
Competence matters because the RIPE environment rewards detailed operational knowledge. A entity who understands routing, resource administration or measurement tools can intervene more credibly in discussions. Status follows when certification or repeated participation signals expertise. Relationships form with trainers and peers. Confidence grows through contact with the institution. Together, these assets can move a person from the periphery toward the group that follows General Meetings and board elections.
That movement is desirable. A registry association should not be governed only by people whose employers have always been able to fund travel and professional development. Training can widen the pool of informed voters and future candidates. It can reduce dependence on inherited expertise. It can help a small operator evaluate budgets and charging schemes with the same seriousness as a large network.
The political value creates a duty of neutrality, not a presumption of wrongdoing. Course locations should be explained by demand, coverage, cost and strategic capacity needs. Registration should use stable rules. Cancellations and waiting lists should be handled consistently. Trainers should not promote candidates or imply that programme continuity depends on a particular election result. Where a board member opens an event, the role should be institutional rather than electoral.
Reporting can expose patterns without turning education into a suspect activity. Annual figures can show courses by country, language, format, entity sector and first-time attendance. The association can compare the distribution with membership geography and operational need. Large unexplained concentrations around politically connected hosts would then become visible. So would persistent neglect of regions that pay fees but rarely receive in-person support.
The RIPE NCC training page makes a useful distinction between in-person courses included in membership and online material available more broadly. That design can reduce scarcity by moving foundational knowledge online. It does not remove the social premium of in-person access. Face-to-face training still creates relationships and local recognition. Integrity controls must therefore address both educational content and the network effects of delivery.
Fellowships carry a stronger obligation
Fellowships are intentionally transformative. The RIPE Fellowship programme describes preparation, coaching, participation during a meeting and follow-up. Its eligibility rules aim at people who live in the service region and work or study in fields connected to Internet operations, governance or community coordination. This is not a casual giveaway. It is an intervention designed to convert potential into sustained participation.
Because the intervention is selective and identity-forming, it carries a stronger patronage risk than a generally available webinar. Selection can determine who receives travel, visibility, mentoring and an accelerated route into community networks. A fellow may be introduced on stage, welcomed into a peer group and invited to continue contributing. Those are legitimate programme features, but they also create a constituency with vivid memories of institutional support.
The selection body should therefore be insulated from current candidates and from officeholders who have endorsed particular applicants. A board may approve the overall purpose and budget, but individual selections should be made under a published mandate by people who are not campaigning and who disclose relevant relationships. If a selector has supervised, employed, nominated or closely collaborated with an applicant, recusal should be recorded.
Scoring should distinguish minimum eligibility from comparative judgment. Completing an introductory course may establish readiness. Residence and field of work may establish scope. Selection among qualified applicants may then consider geographic inclusion, prior access, learning goals and likely contribution. Each factor should be defined well enough that a later reviewer can understand the outcome without reconstructing private impressions.
Appeal need not mean that every disappointed applicant receives a full rehearing. It can provide a channel to correct factual errors, undisclosed conflicts or departures from the announced rules. The reviewer should be separate from the original panel. The existence of review discourages casual exceptions and gives unsuccessful applicants a remedy other than public suspicion.
Public reporting should protect personal circumstances. The association can disclose the number of applications, broad geography, gender balance where lawfully collected, first-time participation, sector, selection stages and recusals without publishing every unsuccessful applicant's name. Recipient names may be public when participation is public and consent is clear, but political analysis should never treat inclusion in a fellowship as proof of allegiance.
Electoral calendars change the risk
A benefit decision made six months after an election does not carry the same immediate risk as one announced during nominations. Timing can turn routine support into campaign currency. If a board member seeking re-election appears at a newly funded course, announces additional vouchers or personally congratulates fellowship recipients shortly before voting, the institutional act can be read as a personal offer even when the programme was planned earlier.
The association should use a campaign-period firewall. Once nominations open, candidates and their campaign supporters should have no role in selecting individual recipients, granting exceptions or announcing discretionary awards. Necessary decisions can continue through staff and independent panels. General information about established programmes can be published in the ordinary way, but communications should avoid candidate-centred credit.
This is not a demand for institutional silence. The RIPE NCC must keep operating during elections. Training calendars cannot be suspended for months, and fellows need timely travel arrangements. A firewall separates administration from promotion. It prevents candidates from using official channels, beneficiary lists or programme events to reinforce electoral claims.
Budget decisions require a related distinction. The elected board may properly propose or approve funding for capacity building. Voters are entitled to assess that record. The danger appears when a candidate moves from defending a policy to cultivating identifiable beneficiaries. A campaign can say, "I support broader training coverage." It should not say, in effect, "Remember who secured your seat."
Communications records can help enforce the line. Benefit notices should come from programme addresses, not candidate accounts. Selection panels should retain conflict declarations. Staff should document late exceptions. Event hosts should receive a short neutrality rule. Candidate access to applicant information should be restricted to what is necessary for board oversight, with aggregate reports preferred over names.
The board's own minutes can record oversight at the right level. They can show the approved budget, objectives, geographic gaps and evaluation findings. They should not become a venue for discussing favoured applicants. The RIPE NCC corporate governance collection presents transparency and a division of responsibilities among members, the board and management as core features. Benefit governance should follow that division.
Incumbency is the central asymmetry
An incumbent candidate enters an election with advantages that no rule can completely remove. The incumbent has a record, institutional familiarity and public visibility. Benefit programmes add another advantage: officeholders can be associated with money and opportunities distributed during their term. Challengers can promise; incumbents can point to delivered support.
That difference is not inherently unfair. Elections are partly judgments on performance. If an incumbent helped establish a successful training strategy, voters may reasonably value that work. The legitimacy problem is not credit for policy. It is control over individual allocation and access to beneficiary relationships that are unavailable to challengers.
The strongest control is functional separation. The board sets strategy and approves resources. Management administers programmes. Independent or mixed panels make selective awards. Staff communicate results. Auditors or designated reviewers test compliance. Candidates receive aggregate information on the same terms. No officeholder should be able to insert a name, accelerate a decision or promise reconsideration.
Informal influence is harder. Staff may anticipate what a powerful board member wants. Selectors may favour applicants linked to a respected incumbent without receiving instructions. Recusal lists and written reasons help, but culture matters. Leadership must state that criticism of the board, support for a challenger or participation in contentious debates is never a negative factor.
The association should also examine host selection. Training delivered through a member organisation can confer prestige and local access. If hosts are repeatedly chosen from networks aligned with incumbents, the pattern can create a regional machine even when individual attendance is open. A transparent call for hosts, objective facility requirements and rotation principles can reduce that risk.
Procurement may overlap with patronage when members provide venues, travel services or educational support. Competitive selection and disclosure of related interests are therefore part of electoral integrity. The relevant benefit is not only what a entity receives. It is also what a host, trainer, sponsor or partner gains from association spending and visibility.
The voter is often an organisation
RIPE NCC voting rights attach to members, while natural persons act for them. This creates another layer between benefit and ballot. A training entity may not be the registered voter. A fellow may not work for a member at all. The person who receives support may nevertheless influence the organisation's authorised representative, prepare its voting recommendation or shape how colleagues understand a candidate.
This indirect path makes disclosure difficult. Publishing a list of beneficiaries and comparing it with votes would be both intrusive and largely useless because ballots are secret. It would also encourage an unfair inference that recipients are captured. The institution should audit the allocation side, not investigate political beliefs.
Organisational voting can amplify certain benefits. A small company may have only one employee who follows governance closely. If that employee receives training and builds trusted ties, their recommendation may determine the member's ballot. A large organisation may have formal government-affairs or network-policy staff who separate course participation from voting. The same benefit therefore has different political weight depending on internal structure.
This variation is not a reason to exclude small organisations. Doing so would entrench large incumbents. It is a reason to make allocation neutral enough that no organisation has to wonder whether voting against an officeholder will reduce future support. The message should be explicit: benefits follow programme rules, not electoral alignment.
Members also need clarity about multiple accounts. RIPE NCC public guidance states that a member with multiple LIR accounts receives one vote, attached to the oldest account. That rule limits one obvious route by which service consumption could multiply formal votes. It does not address influence gained through benefits, but it reinforces an important principle: financial or operational scale should not automatically purchase additional electoral voice within one member.
The same principle should guide education. A large member may send more staff because it has greater need and capacity, but programme statistics should reveal whether repeated attendance by well-resourced organisations crowds out first-time entities. A fair system can reserve some opportunities for newcomers without treating established members as illegitimate. The objective is a wider informed electorate, not a favoured bloc.
Evidence that would reveal a problem
Patronage allegations are easy to make and hard to disprove after the fact. A credible review needs records created before controversy. For universal benefits, the reviewer needs the entitlement rule, capacity data, registration order and exception policy. For selective programmes, the reviewer needs eligibility criteria, scoring guidance, conflict declarations, recusals, anonymised assessments and review outcomes.
Timing data is equally important. The review should compare nominations, campaign periods, award announcements and programme events. A cluster of exceptions close to voting would warrant examination. It would not prove vote buying, but it would show whether normal controls weakened when political incentives were strongest.
Distribution data can identify structural favouritism. Relevant dimensions include country, subregion, organisation type, first-time status, gender where lawfully and voluntarily recorded, language, course subject, host organisation and repeat participation. The comparison should use the eligible population and programme purpose, not a crude expectation that every country receives the same number.
Communications can show whether institutional credit became personal. Reviewers should examine official announcements, candidate statements and event remarks for claims that tie individual access to an officeholder. They should also check whether rejected applicants were offered reconsideration through political contacts rather than the ordinary channel.
Complaints deserve careful treatment. A complaint from an unsuccessful applicant may reflect disappointment, but it may also reveal a hidden exception. The reviewer should test facts without disparaging the complainant. Anonymous or confidential reporting may be necessary where applicants fear losing future opportunities.
Finally, the association should publish the conclusion at a level that supports trust. A report can state how many decisions were tested, how many conflicts were declared, whether criteria were followed and what corrective action occurred. It should not expose personal financial hardship, visa history or private application essays. Electoral auditability and dignity are compatible if the institution audits decision authority rather than political preference.
Designing a neutral benefit constitution
The programme rules should begin with purpose. Training exists to improve operational capability and informed participation. Fellowships exist to lower barriers for new voices with relevant interests. Travel support exists to make attendance possible where cost would otherwise exclude a qualified entity. A clear purpose narrows discretion because selectors must connect each award to an adopted objective.
Eligibility comes next. Rules should specify geography, age where relevant, professional or study connection, prior participation, required learning and any exclusion for repeat awards. Criteria should be available before applications open and remain stable during the round. If a genuine emergency requires a change, the change and reason should be published.
Selection authority must then be defined. Names and affiliations of panel members can be published after decisions if prior publication would invite lobbying. The mandate should bar current board candidates, campaign staff and anyone with a close relationship to an applicant. Recusals should reduce the panel's voting pool rather than being treated as a ceremonial disclosure.
Reasons should be concise but real. A score or note should show why an applicant met the comparative factors. Boilerplate praise does not permit review. The record need not be public at individual level, but an independent reviewer should be able to test consistency.
Appeals should focus on procedural error, conflict and incorrect facts. They should have deadlines and an identified decision maker. A successful appeal may lead to reconsideration, a future-place guarantee or another remedy depending on timing and budget. The remedy should not depend on approaching a board member.
Evaluation should measure more than attendance. It should ask whether entities completed preparation, contributed during or after the event, shared knowledge locally and remained engaged. Those measures help justify the programme without requiring beneficiaries to support institutional leaders. A fellow's critical intervention can be evidence of success, not ingratitude.
What not to do
One bad response is to publish every applicant's identity and score in the name of transparency. That would expose people who disclosed limited funding, career circumstances or barriers to participation. It could deter applicants from smaller organisations and sensitive jurisdictions. Transparency should make the allocator accountable, not make applicants surrender privacy.
Another bad response is to prohibit recipients from participating in elections. Receiving legitimate support does not disqualify a member representative from voting, endorsing or standing as a candidate. Such a ban would create a second-class political category and allow established entities, whose advantages were privately funded, to dominate.
A third mistake is to assume that staff administration eliminates politics. Staff can face pressure, share community loyalties or apply vague criteria inconsistently. Neutrality requires reviewable rules and conflict controls regardless of who signs the decision.
A fourth mistake is to treat every regional correction as favouritism. Programmes may need to spend more in places that have historically received less access. Equity can require unequal distribution. The defence is a published rationale, measurable need and consistent application, not arithmetic uniformity.
A fifth mistake is to let candidates turn integrity rules into attacks on beneficiaries. The recipient is rarely the person who controls the system. Campaigns should challenge allocation design and officeholder conduct, not imply that fellows or trainees sold their votes. Stigmatising support would discourage exactly the participation the programme seeks to build.
The final mistake is to rely on good intentions. A trusted community can resist formal controls because entities know one another and believe overt bribery is implausible. Familiarity is precisely why implicit obligation can be powerful. Rules protect relationships by removing the need to guess whether a favour was political.
A practical campaign-period test
Before each Executive Board election, the RIPE NCC could publish a short benefit-neutrality statement. It would identify programmes operating during the campaign, name the officials responsible for decisions, confirm that candidates cannot intervene and describe the complaint route. This would not require a new bureaucracy. It would make existing separation visible.
The institution could then apply five questions to any award or event. Was the decision made under criteria published before nominations? Did any candidate or close supporter take part? Was an exception granted, and if so, by whom? Did the communication give personal credit to an electoral actor? Can an independent reviewer reconstruct the decision without relying on memory?
If the answer to the second or fourth question is yes, the activity should be reassigned or the communication corrected. If the answer to the first, third or fifth is no, the decision should receive additional review. The purpose is prevention, not scandal management.
After the election, an aggregate report can state how many awards were made, how many recusals occurred, how many exceptions were approved and whether complaints were upheld. The report should cover training seats, fellowship support, travel assistance and other scarce participation opportunities. Routine universal services can be summarised rather than individually audited.
The report should also distinguish the date on which an award was decided from the date on which it was announced. That distinction prevents an ordinary decision made months earlier from looking like a campaign intervention merely because travel confirmation arrived late. Conversely, it stops an institution from disguising a campaign-period exception by attaching it to an older programme round. A reviewer should be able to follow the authority, criteria and timing without learning the recipient's ballot choice.
Hosts and delivery partners require the same attention as individual recipients. A local organisation can gain prestige, access to speakers and influence over invitations when it hosts a course or community event. Host selection should therefore follow geographic need, facilities, accessibility and cost rather than the electoral usefulness of a local ally. Aggregate disclosure of repeat hosts and selection reasons would illuminate this institutional form of benefit without casting suspicion on every partner.
The voting report provides a separate electoral record. For example, the May 2026 General Meeting report states how many members registered and cast ballots and explains the voting methods. Benefit reporting should not attempt to connect named recipients to those ballots. Its function is to show that the electorate was not cultivated through discretionary allocation.
This separation preserves ballot secrecy while improving institutional assurance. The election auditor verifies the ballot and count. The programme reviewer verifies that benefits were allocated neutrally. Members can then judge legitimacy from two independent controls rather than from an impossible demand to reveal how beneficiaries voted.
The case for keeping the benefits
The strongest argument against patronage controls is that they may make programmes cautious, slow and impersonal. Selectors may avoid promising applicants, staff may spend more time documenting decisions, and board members may retreat from useful engagement. Those costs are real. They should be managed by proportionate rules rather than ignored.
The countervailing case is stronger. Training and fellowships are most defensible when entities know that they owe no political debt. A recipient who can criticise the board without fear demonstrates the programme's success. A rejected applicant who receives a credible explanation is less likely to see a closed circle. A challenger who accepts the allocation system as neutral can debate policy rather than insinuate corruption.
Benefits also protect the electorate from inherited inequality. Operators in wealthy organisations can fund their own travel, exams and professional networks. Those advantages are not neutral merely because they are private. Institutional support can give smaller members and newcomers a fairer chance to acquire knowledge and voice. Abolition would leave private patronage untouched while removing the most accountable form of assistance.
The goal is therefore to convert benefits from favours into rights and governed opportunities. Universal access should be treated as entitlement. Scarce access should be allocated by published priority. Selective support should be independent and reviewable. Electoral actors should shape policy at a distance from individual awards.
This design also improves programme quality. Clear purposes produce better selection. Distribution reporting exposes neglected regions. Appeals identify faulty assumptions. Recusals protect panels from reputational damage. Campaign firewalls keep staff out of political disputes. Integrity is not an external burden on capacity building; it is part of competent administration.
The legitimacy threshold
No association can eliminate gratitude, friendship or political affinity. People will support candidates they trust, including leaders associated with programmes that helped them. Governance should not attempt to police those emotions. It should ensure that institutional authority cannot deliberately manufacture them through selective access.
The legitimacy threshold has four parts. First, the benefit must have a stated institutional purpose connected to member service or community capacity. Second, allocation must follow rules that candidates cannot alter for individuals. Third, conflicts and exceptions must be recorded and independently reviewable. Fourth, public reporting must be sufficient to reveal patterns while protecting applicant privacy.
Where those conditions are met, a fellowship recipient's later endorsement is ordinary political participation. A trainee's vote for an incumbent is their own. A host organisation's support may reflect a genuine judgment that the board served the region well. The institution has no reason to apologise for successful programmes.
Where the conditions are absent, even benevolent support can corrode trust. Members may suspect that courses follow allies, that travel support creates a loyal audience or that critics receive fewer chances. Because ballots are secret, those suspicions cannot be resolved by studying votes. They can only be answered by showing that allocation authority was constrained before the election.
The RIPE NCC's educational and fellowship activities are valuable precisely because they shape who can participate. That power should be acknowledged rather than disguised as a collection of apolitical services. Capacity creates voice; voice affects governance. The institution earns legitimacy when it widens both without deciding who should be grateful at the ballot box.
Membership benefits become electoral patronage not when they help people, but when access can be attributed to a person who seeks political return. The cure is neither austerity nor suspicion of recipients. It is a benefit system designed so thoroughly as institutional provision that no candidate can plausibly claim to own the gift.