Who should decide Portland’s next public investments?

01 / Analysis

What a YES or NO vote would do

The existing process includes elected representatives, public hearings, and opportunities for testimony. A NO vote would reject this proposal; it would not legally prohibit future participatory budgeting. A YES vote would create an additional channel for binding decisions; it would not transfer the whole city budget to a public vote. Final ballot filing; full amendment

02 / Analysis

What participatory budgeting actually means

Participatory budgeting, usually shortened to PB, lets residents help decide how a defined pot of public money is spent. The important difference from a survey or public hearing is that residents' final choices control the allocation, within the program's rules.

Under Portland's proposal, the process would follow this sequence:

1. The city establishes the program by ordinance and selects an oversight committee through a process Council determines.
2. Residents submit ideas through meetings and other accessible forums.
3. City staff and resident budget delegates assess and develop a smaller set of feasible projects.
4. Residents select projects through a binding ballot process.
5. City government or its agents implements the winners.
6. Staff and the oversight committee evaluate the program; the City Auditor periodically performs an independent performance audit.

That structure leaves substantial power with public officials and staff. Residents choose among qualified projects, rather than voting to fund anything they want. The screening stage therefore matters as much as the final vote: whoever determines feasibility helps determine the choices residents actually receive. Full amendment, elements 1–6

Illustration, not a promised project: neighbors might propose a safer crossing. Staff would need to determine whether the city can implement it, what it would cost, and whether an available funding source permits that use. It could then compete with other eligible projects. Its selection would not erase engineering, procurement, or implementation work.

03 / Analysis

Read the actual commitment carefully

Source: official full amendment. These distinctions are central to assessing both campaigns' claims.

When would it begin?

The full amendment says funding begins no later than fiscal year 2027–28, and the first PB process begins no later than July 2028. Beginning a process is different from finishing its vote or completing a project.

The final county ballot filing summarizes the program as operational in July 2027. Those descriptions are not identical. The earlier date may describe program setup and funding, but that is an interpretation, not an official clarification obtained for this report. Readers should not assume residents will vote on projects in July 2027. The full text provides the more detailed separation of milestones. Full amendment, opening paragraph; final filing, both pages

The text also uses the term “bi-annual” without defining it. The implementing rules should make the cycle schedule explicit. Regardless of cycle frequency, the funding obligation is annual.

04 / Analysis

Follow the money: three different questions

1. How much would be committed?

The official ballot filing gives preliminary estimates of $16.4 million in 2027–28 and $17 million in 2028–29. Those are estimates in ballot-title language originally filed in August 2025 and carried into the August 2026 county filing. They are not a fixed statutory price or a newly reconciled September 2026 cost estimate. Ballot filing; filing history

The actual floor follows this formula:

Minimum annual PB allocation = 0.02 × prior year's adopted General Fund discretionary ongoing expenses.

For example, an $800 million qualifying base implies $16 million. A $750 million base implies $15 million. These are arithmetic illustrations, not alternative official forecasts.

Portland's adopted-budget overview reports $803.4 million in General Fund discretionary resources, but that summary alone does not establish the amendment's narrower ongoing expenses base. The Budget Office should publish a reconciliation showing exactly which adopted line items count. It would be a mistake to multiply a convenient general-fund number by 2% and present the result as settled. Adopted-budget overview; full amendment

2. Where would the dollars come from?

Portland has many budgets within its overall budget. Money dedicated to water, sewer, transportation, grants, or voter-approved purposes cannot simply be treated as spare cash for any project.

The city's current overview reports $8.55 billion in total budget, but warns that this figure counts internal transactions and overstates spending on actual programs. It reports $4.48 billion in program expenses and describes about 90% of overall funding as restricted. A small percentage of the $8.55 billion headline is therefore a poor affordability test. Adopted-budget overview

Two competing claims both need correction:

• It is too strong to say every PB dollar must come directly from unrestricted General Fund money. The measure allows other sources, and the Budget Office's 2022 memo itself suggested considering capital resources suitable for PB.
• It is equally wrong to imply the entire city budget is freely available. The project must fit the funding source, and using money for PB can displace other eligible work within that source.

The practical question is which lawful, available funds would finance which projects, and what would those dollars otherwise buy? Neither broad assertion answers that. Full amendment; Budget Office memorandum, especially final page

For example, a hypothetical PB-funded parks improvement might fit a restricted parks funding source. That would not turn parks money into unrestricted dollars for any resident-selected service. Nor would it automatically make the improvement additional to everything the parks budget already planned.

3. How much is genuinely additional cost?

The full allocation should not be confused with the cost of administering PB. Most of the allocation can still buy public projects or services. But administration uses resources too.

PB Oregon estimates roughly $1 million a year to run the process. Treat that as an advocate's estimate, not a completed city staffing plan or guaranteed ceiling. The amendment includes administration within program funding; it does not require $16.4 million for projects plus a separately mandated $1 million. Council could choose to allocate more. PB Oregon FAQ, question 17; full amendment

Illustrative annual allocation, using the ballot estimate:

The last two rows are sensitivity examples, not forecasts. They show why a staffing and outreach budget matters. A very cheap process may exclude people; an expensive one leaves fewer dollars for projects. Spending on translation, accessible voting, participant support, and evaluation can have real value. It should still be counted.

Future costs are another issue. A capital project may need maintenance; a temporary service may become politically difficult to end. Portland's amendment does not categorically prohibit recurring commitments. Full project estimates should show startup, ongoing operations, replacement, and the source of future support.

Would it cut essential services?

It creates a real risk of displacement, but does not specify a list of cuts.

If the city uses $16.4 million of flexible money for the program, those dollars cannot simultaneously support Council's alternative spending plan. Some existing services may lose funding. Some PB winners may themselves provide useful services, including work in the same broad categories. The relevant comparison is between the two spending packages, including administration and future obligations.

The Budget Office's October 2022 memo judged bureau reductions a reasonable assumption under the conditions it studied, while explicitly acknowledging uncertainty and possible alternatives. Its figures describe an earlier proposal and earlier budgets, not a 2026 bureau-by-bureau impact assessment for Measure 26-267. 2022 memorandum

A hypothetical $1 million redirected from an existing service into a new service is a real loss to the first service and a gain to the second. Whether Portland is better off depends on their relative value. Calling the entire amount waste ignores the new service; calling it costless ignores what was displaced.

What about Portland's $170 million shortfall?

The June budget announcement described closing a $170 million General Fund gap through a package of actions. It announced shelter-service reductions of $18 million, or 31%, parks reductions of $11.4 million, and about 100 expected layoffs. The 31% figure concerned shelter services, not every homelessness-related expenditure. June budget announcement

There is an important update: Council's July 22 supplemental budget restored 30 positions and provided additional support for public-safety functions. It used $9.96 million of General Fund contingency and $2.25 million from other funds. This does not erase the city's fiscal pressure, but June's projected layoffs and cuts should not be presented as a complete September accounting of what actually occurred. Do not simply subtract 30 from 100 to invent a verified final layoff count. Supplemental ordinance 192207

The $170 million figure describes the gap the city addressed, not an unbalanced $170 million hole still sitting in the adopted budget. Future financial pressure remains a serious consideration. Claims that an eventual forecast surplus will effortlessly cover PB are also uncertain: the favorable future projection cited by PB Oregon comes from a December 2025 forecast, before the full 2026 budget process. PB Oregon FAQ, question 13; linked forecast; budget development record

What happens in a recession?

A percentage-based floor can decline when its qualifying base declines. It is not a legally fixed annual escalator. However, this formula uses the previous year's adopted expenses, so a sudden downturn would not immediately lower the minimum in line with current revenue losses. There is no express emergency waiver in the amendment. That lag is a legitimate concern for opponents. Full amendment

Ten-year totals depend on assumptions. Starting at $16.4 million, a flat allocation totals $164 million over ten years; 3% annual growth totals about $188 million. Those are nominal scenarios, not forecasts. They include administration if it is paid from the same allocation. They represent spending under a different decision process, not necessarily an equal increase in total city spending.

05 / Analysis

Why supporters want a charter guarantee

The campaign's history changes the strongest version of the debate. Supporters are not simply asking whether residents can run a small project-selection exercise. They are asking how to make that power survive changes in political leadership and annual budget priorities.

The supplied timeline describes more than a decade of advocacy, abandoned funding proposals, and eventual use of the initiative process. It is a campaign account, not an independently audited chronology. Its most relevant milestones are:

The strongest YES inference: ordinary budget promises have not produced a durable citywide program, so a voter-approved guarantee may be necessary.

The strongest NO inference: failed starts can also reveal practical capacity and funding problems that a charter obligation alone cannot solve. A credible alternative should explain how it will avoid another abandoned effort.

The regional pilots narrow the uncertainty: Portland is not starting with no local knowledge of PB. They do not remove the uncertainty about this proposal's citywide scale, ongoing funding floor, delivery capacity, or recession performance. Metro's first community-choice round used a $2 million pool in its District 4; its 2026 round awarded $3 million for 23 projects. Those are parks-and-nature investments with restricted purposes, not a test of every provision in Measure 26-267. Metro program; 2026 awards

06 / Analysis

The strongest case for voting YES

These are the best versions of the affirmative arguments, reconstructed from the proposal and evidence. They are not promises about what Portland will achieve.

Residents would gain a power that testimony does not provide

Being allowed to speak and being allowed to decide are different. Someone may participate in hearings for years without knowing whether their contribution changed a budget. A binding project vote offers a visible connection between participation and spending.

For a voter who believes democratic participation is itself a public good, that can justify some process cost even before proving that every winning project is more efficient. The amendment's binding-selection requirement provides a concrete basis for this argument. Full amendment

Best objection: the people who participate might be a small, unrepresentative subset.

Best response: the existing budget process is also unevenly accessible. Compare the reach and influence of both systems, and design PB to recruit beyond habitual participants. More people participating is a plausible benefit; proportional representation still needs to be demonstrated.

Residents can identify needs that centralized processes overlook

People using a street, park, or service can identify small but important problems that are easy to miss in citywide planning. PB offers a structured way to combine that knowledge with staff expertise. Cambridge's 2026 winners included sidewalk and tree work, bike-lane barriers, school supplies, and rat control—ordinary public needs, not merely abstract civic exercises. Cambridge 2026 results

Best objection: city staff already collect needs, and popular projects can crowd out less visible necessities.

Best response: screening and published costs can integrate resident ideas with professional standards. The claim worth testing is whether the combined process finds and serves needs better than existing channels, rather than whether residents or professionals should have exclusive authority.

Broader eligibility can include people affected by decisions

The proposal uses residency, rather than the ordinary registered electorate, as its organizing principle. This can give young people and noncitizen residents a direct role in local spending decisions that affect their lives. Boston and New York already operate PB programs open to residents from age 11, regardless of citizenship or immigration status. Portland's age rules would need clarification. Full amendment; Boston results; NYC citywide process

Best objection: broad eligibility requires different verification and accountability arrangements.

Best response: inclusion and integrity are compatible, but the city must build both. A protected vote is not meaningful if eligible people cannot access it, and an accessible vote is not credible without rules against duplication and manipulation.

Protected funding can prevent a token program

A program with little money or uncertain continuation may struggle to earn people's time and trust. A charter requirement gives communities a reason to invest in repeated cycles, develop skills, and expect a continuing decision-making role.

The campaign policy sheet adds an institutional argument: elected officials are being asked to share their own authority, so leaving the existence and size of PB entirely to their discretion can weaken it. Three experienced PB researchers support the specific amendment for combining a mandate, meaningful resources, and adaptation through evaluation. That is a substantive defense of the charter approach, not merely a claim that PB is popular elsewhere. Policy sheet, pages 2–3; researchers’ letter

Best objection: the same protection can preserve a weak program during a fiscal crisis.

Best response: that is the central tradeoff, not a misunderstanding. A YES voter can reasonably decide that some power should be durably reserved for residents, while expecting operating rules and evaluations to improve its use. But the evidence does not establish that 2% is Portland's uniquely correct amount.

It could build civic skills and longer-term participation

A peer-reviewed New York study linked PB participants to voter-file records and used matching to compare them with similar people outside PB districts. It estimated an 8.4 percentage-point increase in the probability of voting in ordinary elections, with larger estimated effects among some less frequent-voting groups. That is meaningful supportive evidence. It is an observational design, however, not random assignment, and it does not predict an 8.4-point increase in Portland-wide turnout. Johnson, Carlson and Reynolds

Best objection: a successful experience for participants does not prove broad civic transformation.

Best response: assess both. A modest program can still be worthwhile, but the city should measure who participates, who returns, what they learn, and whether benefits extend beyond a small committed group.

07 / Analysis

The strongest case for voting NO

These arguments concern the design and timing of this measure. They do not require rejecting participatory budgeting as a concept.

Portland would make the commitment before knowing the operating plan

Voters know the minimum funding formula but do not yet know the detailed staffing plan, project rules, voting procedures, district allocation approach, or performance targets. Those choices can strongly influence cost and legitimacy.

Best objection: charters normally establish durable principles and leave details to ordinances.

Best response: that is true, but voters may still want a tested operating model before adopting this particular permanent funding floor. The strongest criticism is uncertainty about consequential design choices, not an assertion that the amendment contains no structure. Full amendment

A protected slice makes the remaining budget harder to balance

The city must meet the PB obligation even when leaders judge other needs more urgent. In a downturn, services outside protected allocations may shoulder a larger share of adjustments. The previous-year formula adds a timing constraint.

Best objection: elected officials should not be the only people defining which needs are urgent.

Best response: that is precisely the disagreement. A NO voter can support more public influence while preferring elected officials to remain accountable for tradeoffs across the whole budget, including emergency response and ongoing obligations.

Participation could reproduce existing inequalities

Organized groups have time, networks, and campaign skills. Residents facing long work hours, caregiving duties, language barriers, or unstable housing may struggle to participate. Those differences affect idea submission and project development as well as voting.

A Cook County study analyzed 1,446 survey respondents and found greater interest in PB among people with higher incomes and more education. It measured stated interest, not participation in a Portland program, but it challenges any assumption that simply opening a ballot automatically broadens representation. Doherty and colleagues

Best objection: deliberate outreach and compensation can change those patterns.

Best response: yes, and those efforts require resources, credible targets, and evidence of results. A NO voter may want that model demonstrated first. There is no evidence here establishing that capture by any named Portland organization is inevitable.

Successful votes do not ensure successful delivery

Projects can run into higher costs, staffing limits, site restrictions, or difficult procurement. Delays can undermine the trust PB is intended to build. Seattle's 2023 vote did not lead to authorization for implementation until the 2025–26 budget, and implementation work continued beyond that. Seattle program timeline

Best objection: ordinary city projects face the same problems.

Best response: correct. The proper test is comparative delivery quality, not whether PB ever experiences a delay. Portland should require realistic estimates, accountable owners, and public progress reporting before asking residents to invest their hopes in a winning ballot item.

A pilot could preserve the opportunity to learn

City Club of Portland's March 2025 study recommends a limited pilot, followed by expansion if justified. It recognizes potential benefits while questioning the evidence for a permanent citywide commitment of the kind being discussed. Its recommendations favor explicit goals and evaluation, followed by consideration of a district-based approach. City Club report, pages 53–59

Best objection: another pilot or study can become an excuse for indefinite delay.

Best response: opponents should attach a concrete alternative: funding source, start date, responsible department, evaluation criteria, and a public decision on expansion. A NO vote alone supplies none of those things.

08 / Analysis

What other places actually tell us

The most useful comparisons separate program design, participation, project selection, delivery, and measured outcomes. A city announcing winners has established selection, not necessarily completion or value for money.

The presentation makes the service choices concrete

The campaign's 59-slide presentation illustrates a wide range of possible investments: safe crossings, fire equipment, accessible pools, food access, childcare, mental-health response, public restrooms, housing support, and youth programs. These are examples from other places, not promises of eligible or winning Portland projects. Campaign presentation, slides 19–44

Two examples can be checked against city announcements. Somerville's 2023 selection included $55,000 for community food access and $250,000 to accelerate bike lanes. Seattle's 2023 winners included $7.2 million for public-restroom staffing and maintenance and $2 million for a crisis-response team. These records establish selection and intended uses, not independently measured benefits or completion of every project. Somerville results; Seattle results

This matters to both sides: PB can fund useful services, so its entire budget is not automatically lost to services. But selecting a service through PB does not establish that it meets the city's most urgent need, costs less, or avoids displacing another service.

The research supports potential, not a universal promise

Some studies find substantive gains. Sonia Gonçalves's study of Brazilian municipalities during 1990–2004 found PB adoption associated with more spending on sanitation and health and lower infant mortality. It used longitudinal municipal data and robustness checks. That is stronger evidence than testimonials, but it remains evidence from a very different institutional and service context. It should not be converted into a Portland health or savings forecast. World Development study

The overall evidence has limitations. A systematic review covering 139 papers from 1989–2019 found a diverse field and mixed evidence concerning trust, transparency, and accountability. This undercuts claims that widespread adoption establishes universal success. It also does not establish universal failure. Bartocci and colleagues

Design matters. Urban Institute guidance recommends sufficient funding, intentionally inclusive outreach, paying people for their time, multiple participation methods, follow-up, and measurement. It also recommends securing continuity through legislation after a successful pilot. These recommendations support parts of both sides' strongest arguments. Guidance is not proof that Portland's exact design will work. Urban Institute

Our assessment: the evidence is reasonably persuasive that well-run PB can create meaningful participation and select useful projects. It is less decisive about durable citywide trust, net fiscal efficiency, or whether a permanent minimum linked to unrestricted spending is superior to a pilot or targeted program in Portland.

What the researchers' open letter contributes

An August 30, 2026 letter by Brian Wampler, Stephanie McNulty, and Michael Touchton, supplied by Your 2 Cents, directly assesses Measure 26-267. The authors co-wrote Participatory Budgeting in Global Perspective (2021). Their letter supports the measure; it is an expert assessment taking a position, rather than a neutral fiscal review or a new Portland impact study. Read the two-page letter

Their strongest points are specific to the design:

• Durability with adaptation. They favor a charter mandate that preserves room to refine implementation. This fairly distinguishes changing process rules from eliminating the protected funding floor.
• Enough money to matter. They argue that small allocations can make participation unrewarding, and identify at least $1 million per 100,000 residents as a useful benchmark.
• Evaluation as part of the program. They credit the explicit requirement to evaluate and improve participation and outcomes.
• Local civic capacity. They see Portland's civic organizations and social movements as an asset for recruitment and implementation.

The funding benchmark has a traceable basis: the Participatory Budgeting Project's 2017 scoping toolkit, page 11, recommends that starting scale for a city or district. It is practitioner guidance from a PB-promoting organization, not an experimentally established threshold or a demonstration that Portland's exact 2% formula is optimal. $1 million per 100,000 residents equals $10 per resident; evaluating a Portland comparison requires consistent population, year, and treatment of administrative costs. Scoping toolkit

The letter strengthens the YES case that a mandate and a substantial budget can be features of a successful design. It does not supply a Portland staffing budget, identify the spending displaced, estimate likely turnout, or compare this measure against a fully specified pilot. Its judgment that Portland's civic networks are favorable conditions remains a prediction. Those same networks could broaden access through outreach or dominate it if safeguards are weak; their effects should be measured.

How this changes our assessment: a charter mandate should not be dismissed as inherently incompatible with learning. This amendment explicitly requires learning and permits process changes. The unresolved tradeoff is whether that flexibility is enough when the annual funding obligation itself cannot be suspended by ordinary program rulemaking.

Read City Club critically, too

The 2025 report is an important independent local resource, not an official 2026 ballot ruling. It predates the July 2025 filing of this exact amendment. Some descriptions also deserve correction: it calls Cleveland's rejection decisive in language stronger than the roughly 51–49 result supports, and its discussion of NYC should be checked against the distinction between a charter-created process and a guaranteed percentage of spending. Its central pilot recommendation remains relevant; its historical details should not be repeated uncritically. City Club report, page 39; Cleveland results; NYC charter background

09 / Analysis

Checking the campaign claims

The YES website is Your 2 Cents Portland. The NO website is Protect Our City Services. Both are advocacy sources. The table below paraphrases claims rather than reproducing campaign language. The assessment rests on the underlying records.

Two further clarifications matter. First, the Charter Commission's earlier PB proposal did not meet the threshold for direct ballot referral, but it was recommended to Council. Describing that history as a simple rejection omits that recommendation. The city formally received the recommendations in January 2023. Council report 66-2023

Second, financial estimates changing between years does not by itself prove that the formula is incoherent. Forecasts change. The legitimate question is whether the city can publish a transparent calculation from the actual adopted budget using the amendment's precise terms. NO FAQ; full amendment

10 / Analysis

Who benefits, who bears the risk?

The outcome cannot be assigned in advance to a political party, neighborhood, or organization.

Potential beneficiaries include residents whose needs are overlooked, people who cannot participate in ordinary elections, neighborhoods with well-developed projects, and anyone served by successful winners. Civic groups may help more people participate and may also become project implementers.

People exposed to costs include users of services that lose alternative funding, staff asked to administer the process, and residents whose preferred projects do not make the ballot or win. Participants also give time and effort; disappointment is costly if projects stall.

Those categories overlap. A person could gain influence through PB and lose hours at a service they use. A neighborhood could win a new amenity but face reduced maintenance elsewhere. That is why participation counts and project announcements alone cannot establish net benefits.

An organization's endorsement is evidence of its position, not evidence that it will receive funding or that its argument is wrong. The measure does not award contracts to named campaign backers. Conversely, a future public grant should be judged by selection rules, conflicts, performance, and public benefit—not by the assumption that a community organization is automatically trustworthy. Full amendment

The two coalitions are broader than a simple labor-versus-business story. The YES site lists labor, civil-rights, political, community, and business supporters. The NO coalition page lists the Northwest Oregon Labor Council, Portland Fire Fighters' Association, and several elected officials. These are campaign-published endorsements, not a completed independent verification of each organization's authorization. YES endorsements; NO coalition

11 / Analysis

What would make the proposal easier to judge?

The following information could materially change the assessment. These are research questions and suggested implementation standards, not additional requirements already contained in the amendment.

A useful evaluation would compare participation with existing channels; report reach by district and relevant demographics using privacy-protecting methods; publish rejection reasons and conflict disclosures; track both process and project costs; and measure delivery and service results. It should include nonparticipants and affected service users, not only satisfied PB participants.

12 / Analysis

A fair way to make the voting decision

A YES vote is most defensible if you place substantial value on binding resident power, think durable funding is necessary to make it meaningful, and are willing to accept budget constraints and unresolved operating details while the city builds the program.

A NO vote is most defensible if you want PB tested before this permanent financial commitment, give greater weight to flexibility and existing service obligations, and prefer to pursue a specific, funded alternative through the ordinary budget process.

Neither vote is a reliable proxy for caring about neighborhoods, trusting ordinary residents, or supporting essential services.

The evidence establishes a real opportunity and a real constraint. The opportunity is to give more residents direct authority over public investments. The constraint is a continuing claim on scarce resources, with implementation choices still ahead. The unresolved question is whether Portland's gains from that shift in power will justify what it costs and displaces. That is the substantive choice voters are being asked to make.

13 / Analysis

Read the supplied campaign materials

All six items below informed this update. The five PDFs are preserved as supplied. The presentation remains hosted by the campaign. Campaign attribution identifies the source of a claim; it does not imply this analysis endorses the campaign. The official filed amendment controls when a summary differs from it.

The fact sheet estimates roughly 150,000 residents cannot vote in ordinary elections, and the deck describes stronger support among certain demographic groups. No population calculation or poll methodology was supplied in this packet, so those numbers and subgroup claims are not treated here as verified measurements. Broader resident eligibility is established by the amendment itself. Policy sheet, page 2; presentation, slides 17–18

The presentation's global comparisons are a useful research lead, not a standardized ranking. They mix voters, project votes, participants, total populations, annual budgets, and multiyear programs. A chart showing both spending and participation growing cannot by itself isolate whether more funding caused higher turnout. We use the separately sourced city records above rather than repeat every slide statistic as fact. Presentation, slides 6 and 52–59

14 / Analysis

Sources and scope

The September 17 baseline reviewed both requested campaign sites and their substantive policy pages; the official measure text and final county filing; Portland budget materials including the July supplemental ordinance; local history; official PB records from other jurisdictions; and research with different methods and findings. The September 20 update adds all five supplied PDFs, the full 59-slide presentation, and targeted checks of the new claims. Links appear beside the claims they support.

Official text is used to establish requirements. Campaign pages establish campaign positions. Government project announcements establish reported selections and process facts, not independent impact evaluations. Academic and civic research inform likely benefits and risks; their findings are not treated as guarantees for Portland.

This is a substantial desk review, not an exhaustive or preregistered systematic review. It does not contain original interviews, a legal opinion on unresolved implementation questions, or a current transaction-level campaign-finance audit. No campaign was asked to approve the analysis. The supplied campaign documents are treated as source material; private correspondence is not reproduced or treated as an interview. The September 20 update is not a complete new audit of every earlier source. Research notes record source coverage and outstanding verification needs.