Canada regular EI beneficiaries, June 2026

Framing and exact resolver: this targets Statistics Canada Table 14-10-0011-01, Canada, regular Employment Insurance beneficiaries, monthly, seasonally adjusted, June 2026, first print, converted from persons to thousands. The ledger table description mentions The Daily release for May 2026, but the catalog slug, dataPointId, expected release window, and question period identify June 2026, so I keep the forecast tied to the June 2026 target.

Resolution timing: the Statistics Canada 2026 release schedule and the April EI release notice place the EI sequence on July 23, 2026 for May 2026 and August 20, 2026 for June 2026 within the registered August 19 to August 27 window. I therefore use 2026-08-20 as the first-print resolution date rather than inferring solely from cadence.

Reference class and base rate: for a two-month-ahead level forecast of a seasonally adjusted EI beneficiary count, the base rate is persistence from the latest available level with recent official monthly changes. Same-variant anchors are all Canada regular EI beneficiaries, seasonally adjusted, from the same Statistics Canada EI series, not gross claims, LFS unemployment, or unadjusted counts.

Prior/update/interval: persistence prior starts at April 2026 = 544.44 thousand. Recent official EI changes used for dispersion are January -17.0 thousand, February -8.7 thousand, March +2.3 thousand, and April about -2.7 thousand from -0.5% of roughly 547.2 thousand; mean change = -6.5 thousand and sample sigma = 8.3 thousand. Base two-month trend prior is 544.4 - 2*6.5 = 531.4 thousand. I add +5.6 thousand for March-April stabilization and EI benefit duration lag, then subtract 2.0 thousand for the strong May LFS signal lowering expected inflows, for a net +3.6 thousand adjustment and point = 535.0 thousand. The short four-change volatility sample is narrow but same-variant and recent, so I use it as a base and widen for horizon. The one-month 80% half-width is 1.28*sigma = 1.28*8.3 = 10.6 thousand; because June is two unpublished EI months beyond April, I widen to about 16.0 thousand, within 1.5x the one-month half-width, so interval = 535.0 +/- 16.0 = [519.0, 551.0].

Upside risk: tariff-sensitive layoffs, administrative backlogs, or slower exits from regular benefits would land above the interval if the first-print June 2026 Table 14-10-0011-01 value exceeds 551 thousand. Downside risk: a broad job-finding improvement after the May LFS rebound would land below the interval if the first-print June 2026 value falls below 519 thousand. Outside the interval would be most plausible if the May LFS employment gain carries directly into EI exits or if a sudden sectoral shock reverses it before the June EI reference week.

Review disposition: accepted the resolver critique by naming the August 20, 2026 official release-calendar basis, accepted the update critique by separating stabilization, EI lag, and LFS components, and accepted the interval-sample note by flagging the four-change volatility base and horizon widening.

The resolver is the Canada, seasonally adjusted regular-beneficiary total in Statistics Canada table 14-10-0011-01, converted from persons to thousands. This uses the same SA variant for every anchor; it resolves only to the first June 2026 print, not later table revisions.

Base rate/reference class: the 2025-June through 2026-April SA level sequence is elevated but its recent movement has been mixed, with the late-2025 high followed by a January-February decline and a partial March rebound. The outside-view forecast therefore begins near the latest level rather than extrapolating either the December peak or the one-month April fall.

From the 10 successive changes in the official June 2025-April 2026 history (7.39, 0.18, -1.00, 7.21, 7.24, -1.10, -16.78, -8.73, 5.89, -3.56 thousand), the sample standard deviation is 7.93 thousand; 1.28 × 7.93 = 10.15 thousand for an 80% half-width.

Prior/update/interval: persistence prior is April's 544.44 thousand; the historical sample is the 10 official successive SA changes from June 2025 through April 2026; adjustments are a small mean-reversion offset after the recent January-February drop, tempered by April's -0.5% decline and two unobserved monthly steps; sigma = 7.93, so the interval method uses 1.28*sigma = 10.15 thousand, rounded to 10.2; final implied bounds are 534.3 to 554.7 thousand around 544.5.

upside risk: renewed job loss or unusually large new claims could push beneficiaries above 554.7 thousand. downside risk: a stronger return-to-work flow or benefit exhaustion could push the result below 534.3 thousand. A monthly move larger than the recent historical dispersion would land outside the interval.

The resolver is Statistics Canada table 14-10-0011-01: Canada, regular benefits, both sexes, age 15 years and over, seasonally adjusted. Every level anchor below uses that same SA regular-beneficiary variant. The registered target specifies the first June print on 2026-08-20; the official Daily schedule consulted in this run is a rolling two-week calendar, while the prior EI release explicitly announced the May release for 2026-07-23.

Base rate: the relevant reference class is the latest four monthly changes in the same SA level series: -8.60, -8.67, -2.91, and -3.00 thousand. The near-term base case is continued but slower decline after the sharp January-February falls; stronger May employment and a lower unemployment rate modestly reinforce a lower June EI-beneficiary level.

Prior/update/interval: persistence prior uses the mean of the latest two changes, (-2.91 - 3.00)/2 = -2.955 thousand per month; two-step persistence from April gives 544.44 - 2×2.955 = 538.53. I apply a -1.03 thousand labour-market update from May's 88,000 employment gain and 0.3-point unemployment-rate fall, giving 537.50. Using the four fetched successive changes, sample sigma = 3.27 thousand; 1.28×sigma = 4.19 thousand, rounded to a 4.20-thousand 80% half-width. Final implied bounds are 537.50 - 4.20 = 533.30 and 537.50 + 4.20 = 541.70 thousand.

Upside risk: renewed layoffs or a reversal of May's employment gain would raise regular beneficiaries above the interval. Downside risk: unusually rapid job finding or benefit exits would push the print below the interval. An administrative or eligibility-policy change affecting receipt counts would be outside the interval because it is not captured by recent SA month-to-month dispersion.

The resolver is Statistics Canada table 14-10-0011-01: Canada, regular-beneficiary detail total, sex total, age-group total, seasonally adjusted. All anchors below use that same variant; values are persons converted to thousands by 0.001, and only the first June print counts.

Base rate/reference class: the near-term reference class is the four latest matching seasonally adjusted Canada totals, 554.4, 545.7, 548.0, and 544.44 thousand. The April level and the May labour-market improvement favour a small decline rather than a return to the November 2025 peak of 569.0 thousand.

Prior/update/interval: persistence prior is April's 544.44 thousand, using the January-April 2026 matching-variant historical sample (554.4, 545.7, 548.0, 544.44). Successive changes are -8.7, +2.3, and -3.56 thousand; sample sigma = 5.5 thousand, so 1.28*sigma = 7.0 thousand. I apply a -3.4 thousand combined adjustment for recent level/momentum (slightly negative), the May employment and unemployment improvement (negative), and offsetting administrative/eligibility uncertainty (positive), giving 541.0 thousand and implied 80% bounds of 534.0 to 548.0 thousand.

Counter-consideration: upside risk is a renewed rise in layoffs or eligibility-driven inflows, which would land above the interval. Downside risk is a stronger-than-May employment recovery and faster exits from benefits, which would land below the interval. A material administrative-policy change would also put the first print outside the interval.

Derived run: the pointwise median of the CDFs of 3 independent thesis.analyst fast rollouts on this target — median prediction sampling per Turtel et al. 2025 (arXiv:2505.17989). No new model call; this run is a deterministic aggregate of the recorded rollouts at 2026-07-10T15:38:24Z, 2026-07-10T15:42:29Z, 2026-07-10T15:46:34Z. Drivers and resolver fields mirror the rollout closest to the median.

Median CDF quantiles: q10 = 533.9, q50 = 541.0, q90 = 548.1. Constituent points [544.5, 537.5, 541.0] with 80% widths [20.4, 8.4, 14.0]; the median interval inherits the central rollout mass rather than averaging tails.

Framing and exact resolver: this is Statistics Canada Table 14-10-0011-01, Canada, seasonally adjusted regular Employment Insurance beneficiaries, beneficiary detail total, sex total, age group total, June 2026 first print, converted from persons to thousands. The ledger sourceBinding table label says May 2026, but the dataPointId, period, resolutionDate, and resolutionRule correctly target June 2026, so I keep the forecast tied to the June target.

Reference class and base rate: for the same SA Canada regular-beneficiaries variant, the official January-April 2026 first-print values are 554.83, 542.11, 548.00, and 544.44 thousand; a naive base rate is the recent level around 547.35 thousand, with month-to-month changes of -12.72, +5.89, and -3.56 thousand.

Prior/update/interval: persistence prior = April 2026 first print of 544.44 thousand, historical sample = official Jan-Apr 2026 first prints for the exact SA Canada total-beneficiary variant, adjustment components = -4 thousand for firm May/June labour-market conditions, -1 thousand for continued exhaustion/return-to-work flow, +0 thousand for policy because no target-specific EI rule change was identified. Successive changes are -12.72, +5.89, -3.56 thousand, so sample sigma = 9.3 and 1.28*sigma = 11.9 thousand. I widen because the target is two unreleased EI months ahead and the current/previous month are revision-prone before the first print: ladder-implied 80% width is 557.5 - 522.0 = 35.5 thousand, half-width 17.75 thousand, which is about 1.49x the 1.28*sigma half-width.

Ladder: P(X <= 515) = 0.05; P(X <= 520) = 0.08; P(X <= 525) = 0.13; P(X <= 530) = 0.22; P(X <= 535) = 0.36; P(X <= 540) = 0.52; P(X <= 545) = 0.66; P(X <= 550) = 0.78; P(X <= 555) = 0.87; P(X <= 560) = 0.93; P(X <= 565) = 0.97. Linear interpolation gives p10 = 522.0, p50 = 539.4, and p90 = 557.5 thousand.

Counter-considerations: upside risk is a delayed EI response to early-2026 unemployment that keeps beneficiaries near or above April and would land above the interval if June prints above 557.5 thousand. Downside risk is that strong May hiring plus June job finding quickly pulls recipients off regular benefits and would land below the interval if June prints below 522.0 thousand. Outside the interval would most likely require an abrupt policy/administrative processing shift, a major layoff wave, or much larger-than-normal first-print revision behavior.

Review disposition: accepted the reviewer suggestion to ground the labour-market context in the official Statistics Canada Labour Force Survey release rather than the secondary WSJ citation, kept the ledger discrepancy visible, and did not add a longer volatility sample because the draft's interval arithmetic is explicitly tied to the same first-print 2026 reference class and widened for the two-month forecast horizon.

Framing and exact resolver: the target is Statistics Canada Table 14-10-0011-01, Canada, regular Employment Insurance beneficiaries, monthly, seasonally adjusted, beneficiary detail total, sex total, age group total, June 2026 first print, converted from persons to thousands. The sourceBinding table text says May 2026 in one place, but the slug, dataPointId, period, and resolution rule all identify June 2026, so I keep the forecast tied to the June target.

Release-date check: the Statistics Canada major-release calendar says data are released at the time of release on the Statistics Canada website and in The Daily, and the ledger resolution date is 2026-08-20. The same official calendar verifies adjacent August 20, 2026 major releases, while the April 2026 EI Daily notice says May 2026 EI data will be released July 23, supporting the monthly publication sequence into the August window for June EI.

Base rate/reference class: for this level series, the base rate is a persistence prior around the latest official EI stock, because monthly SA regular-beneficiary changes are usually single-digit thousands unless labour-market conditions shift sharply. The same SA variant is used for every anchor: Canada, regular beneficiaries, seasonally adjusted.

Prior/update/interval: persistence model prior = April 2026 level 544.44 thousand; historical sample = recent official first-print level changes in thousands from Nov-Dec, Dec-Jan, Jan-Feb, Feb-Mar, Mar-Apr, using 569.0, 567.8, 550.8, 542.1, 548.0, 544.44 where Dec/Jan are derived from official stated monthly declines; adjustment components = -4.0 thousand for better May/June labour-market conditions, -3.0 thousand for lagged exits after strong May employment, -1.4 thousand for no new policy expansion; point = 544.44 - 8.44 = 536.0 thousand. The one-month change dispersion is sigma = 7.5 thousand; for a two-month-ahead June target I scale by sqrt(2), so sigma = 10.6 thousand and the 80% half-width is 1.28*sigma = 13.6 thousand, giving 536.0 +/- 13.6 = [522.4, 549.6].

Counter-considerations: upside risk is that EI beneficiary stocks lag the June unemployment improvement and remain near the March-April range, which would land above the point but mostly inside the interval. Downside risk is that May and June job gains quickly reduce new and continuing beneficiaries, pulling the level into the low 520s. Outside the interval would require either a renewed layoffs shock pushing June above 549.6 thousand or an unusually fast beneficiary exit wave pushing June below 522.4 thousand.

Framing and exact resolver: the target is Statistics Canada Table 14-10-0011-01, Canada, regular Employment Insurance beneficiaries, seasonally adjusted, beneficiary detail total, sex total, age group total, June 2026 first print, converted from persons to thousands. I keep the ledger date and resolver; the sourceBinding table label says May 2026 while the dataPointId and resolutionRule say June 2026, so I treat that as a label discrepancy rather than changing the target.

Reference class/base rate: for this seasonally adjusted level series, I use recent official first-print Canada EI levels as the base rate, not the catalog forecast. The January-April sequence was 554.83, 542.11, 548.00, and 544.44 thousand; persistence from the latest official print is therefore a mid-540s baseline.

Variant check: all anchors above are the same variant as the resolver, namely regular EI beneficiaries for Canada, seasonally adjusted, from Table 14-10-0011-01, not claims, not unadjusted data, and not LFS unemployment.

Prior/update/interval: persistence prior = April 2026 level 544.44 thousand from official first-print history; historical sample = Jan-Apr 2026 official levels 554.83, 542.11, 548.00, 544.44; adjustment components = -2.0 thousand for mild downward momentum from Jan-Apr, -2.5 thousand for June LFS unemployment easing to 6.5%, and about 0.0 thousand for no identified EI policy shock; raw point = 544.44 - 2.0 - 2.5 = 539.94, rounded to 540.0. Successive monthly changes are -12.72, +5.89, and -3.56 thousand, so sample sigma = 9.3. A one-month 80% half-width is 1.28*sigma = 1.28*9.3 = 11.9; because the June target has two unpublished EI reference months after April, I widen to about 17.0 thousand, giving 540.0 +/- 17.0 = [523.0, 557.0].

Counter-consideration: upside risk is that May and June EI counts respond with a lag to earlier labour-market softness or benefit-duration effects, which would land above the interval if the first print exceeds 557 thousand. Downside risk is a faster unwind in claims after the June LFS improvement, which would land below the interval if the first print is under 523 thousand. Outside the interval would likely require either a broad administrative/policy shift or a much sharper labour-market move than the recent official series shows.

Framing and exact resolver: this forecast targets Statistics Canada Table 14-10-0011-01, Canada regular Employment Insurance beneficiaries, monthly, seasonally adjusted, beneficiary detail total, sex total, age group total, June 2026 first print, in thousands. I keep the ledger first-print rule and the resolutionDate 2026-08-20; the public release-schedule page states releases are published in The Daily at 8:30 a.m. Eastern and the latest EI Daily release states the next EI reference month schedule, so I treat the ledger date as the registered target schedule rather than inferring it from cadence.

Reference class and base rate: for this same SA Canada EI variant, the recent first-print level reference class is 569.11 thousand in December 2025, 554.83 in January 2026, 542.11 in February, 548.00 in March, and 544.44 in April. A persistence base rate from the latest first print is therefore 544.44 thousand before May-June updates.

Level, momentum, one-off, and policy mechanisms: the level is still high versus early 2025 on year-over-year comparisons, but the January-February drop and the April -0.5% print show the late-2025 climb has stalled. Momentum is slightly downward. The June LFS improvement lowers the forecast modestly, while EI lag and eligibility/exhaustion mechanics keep the adjustment smaller than the employment headline. I apply about -7.9 thousand over May-June from the April base.

Prior/update/interval: persistence prior = April first print 544.44 thousand from Table 14-10-0011-01; historical sample = first-print Dec 2025-Apr 2026 values 569.11, 554.83, 542.11, 548.00, 544.44 thousand; adjustment components = -4.0 thousand continuation of stalled/downward EI momentum, -3.9 thousand from improved June labour conditions, 0.0 thousand explicit policy change adjustment; point = 544.44 - 7.94 = 536.50 thousand. Monthly successive changes are -14.28, -12.72, +5.89, -3.56 thousand, so sigma = 9.3 thousand. For a two-unreleased-month horizon, 80% half-width = 1.28 * sigma * sqrt(2) = 1.28 * 9.3 * 1.414 = 16.9 thousand, giving 536.5 +/- 16.9 = 519.6 to 553.4 thousand.

Counter-consideration: upside risk is a higher May-June beneficiary count if layoffs in manufacturing, construction, or tariff-sensitive goods sectors continued despite headline employment gains; that could land above the interval if both May and June add roughly 9 thousand or more. Downside risk is faster exhaustion or re-employment after the earlier 2025 spike; two strong negative EI prints like January-February would land below the interval. Outside the interval would most likely require a broad administrative or eligibility shock, a large revision to the SA setup before first print, or a labour-market move not visible in the June LFS headline.

Derived run: the pointwise median of the CDFs of 3 independent thesis.analyst fast rollouts on this target — median prediction sampling per Turtel et al. 2025 (arXiv:2505.17989). No new model call; this run is a deterministic aggregate of the recorded rollouts at 2026-07-10T16:23:42Z, 2026-07-10T16:35:20Z, 2026-07-10T16:46:34Z. Drivers and resolver fields mirror the rollout closest to the median.

Median CDF quantiles: q10 = 522.2, q50 = 536.5, q90 = 553.5. Constituent points [536.0, 540, 536.5] with 80% widths [27.2, 34, 33.8]; the median interval inherits the central rollout mass rather than averaging tails.

The target is the Canada total for regular-benefit recipients, seasonally adjusted, both sexes and age 15 years and over, in Statistics Canada Table 14-10-0011-01. All anchors below use this same variant. The table reports persons, converted to thousands by multiplying by 0.001.

The recent reference class has a roughly flat-to-declining level after the late-2025 peak: 569.11 thousand in December, 542.11 thousand in February, 548.00 thousand in March, and 544.44 thousand in April. This supplies the base rate: short-horizon persistence around the latest level, with monthly noise near ten thousand.

Level is 544.44 thousand in April. Momentum is mildly negative across the latest observations. The May and June employment gains are a downside adjustment because stronger employment should reduce entries and speed exits, although EI administrative lags weaken the immediate effect. Temporary tariff-related EI measures lasting through October 10 are an upside adjustment because easier access and longer entitlement can sustain beneficiaries.

Prior/update/interval: persistence model prior = April's 544.44 thousand; historical sample = November 565.96, December 569.11, inferred January 550.81 from the published February decline, February 542.11, March 548.00, and April 544.44. Successive changes are +3.15, -18.30, -8.70, +5.89, and -3.56 thousand; sample sigma = 9.68 thousand. Adjustments are -4.0 thousand for stronger May-June employment and +1.5 thousand for temporary EI eligibility/entitlement measures, giving 544.44 - 4.0 + 1.5 = 541.94, rounded to 542.0. The 80% half-width is 1.28*9.68 = 12.39 thousand, producing 542.0 ± 12.4 = [529.6, 554.4].

Upside risk would come from tariff-related layoffs, broader uptake under temporary EI measures, or delayed exits and would land above 554.4 thousand. Downside risk would come from the May-June employment improvement rapidly reducing claims or accelerating returns to work and would land below 529.6 thousand; either outcome is outside the interval.

The target is the first-print Canada total in Statistics Canada Table 14-10-0011-01: regular benefits, seasonally adjusted, both sexes, age 15 years and over. All anchors use this same variant and are converted from persons to thousands. The ledger carries 2026-08-20, but Statistics Canada's official 2026 release-date PDF currently schedules Employment Insurance for June 2026 on 2026-08-19; this is a concrete one-day ledger discrepancy, so the forecast remains tied to the registered target while flagging it.

The reference class/base rate is short-horizon persistence in this slowly moving administrative level series. Recent first-print monthly changes were -14.110, -13.000, +6.000, and -3.560 thousand; their mean was -6.168 thousand, although the two latest changes show much less deterioration than January and February.

Level and momentum point mildly downward from April's 544.440 thousand. The May employment rebound and lower unemployment rate argue against extrapolating the early-2026 declines, while the lag between joblessness and benefit receipt means they do not imply an immediate sharp fall. Temporary EI measures are a policy-mechanism support to beneficiary counts.

Prior/update/interval: persistence prior = 544.440 thousand using the December 2025-April 2026 first-print sample; adjustment components = -2.0 thousand for residual claims momentum and -0.94 thousand for the net May-June labour-market and policy mix, giving 544.440 - 2.000 - 0.940 = 541.500. From successive changes [-14.110, -13.000, 6.000, -3.560], sample sigma = 9.39 thousand; the normal 80% half-width is roughly 1.28*sigma = 1.28*9.39 = 12.02 thousand, rounded to 12.0, yielding 541.5 ± 12.0 = [529.5, 553.5].

Upside risk is a delayed rise in eligible claims following earlier job losses or stronger take-up under temporary EI measures, which could land above 553.5 thousand. Downside risk is faster job finding, benefit exhaustion, or the May employment rebound passing quickly into claims, which could land below 529.5 thousand; either would be outside the interval.

The resolver is Statistics Canada Table 14-10-0011-01: Canada, regular benefits, both sexes, age 15 years and over, seasonally adjusted, first June 2026 print, converted from persons to thousands. All EI anchors below use that same first-print variant.

The ledger appears to contain a concrete one-day scheduling discrepancy: Statistics Canada's calendar says August 19, while the target contract says August 20. I retain the contract's resolutionDate and exact dataPointId, but resolution operations should monitor the official August 19 release.

The reference class is the four overlapping two-month first-print changes from the recent official series level; its base rate is continued softening from the late-2025 peak, tempered by the May employment rebound and lower unemployment.

Prior/update/interval: a two-month persistence model uses the first-print historical sample December 2025 through April 2026. Successive two-month changes are -23.25, -6.83, and +2.33 thousand; their mean is -9.25 and sample sigma = 12.96 thousand. The persistence prior is 544.44 - 9.25 = 535.19. Add +4.0 thousand for the May labour-market improvement partly arresting the decline, with no separate policy adjustment, giving 539.19, rounded to 539.2. The 80% half-width is 1.28*sigma = 1.28*12.96 = 16.59 thousand, producing 539.2 ± 16.6 = 522.6 to 555.8 thousand.

Upside risk is renewed layoffs, slower benefit exits, or delayed transmission of the May hiring improvement; a return toward the late-2025 beneficiary peak would land above the interval. Downside risk is sustained job finding, benefit exhaustion, or another broad employment gain; a two-month fall exceeding the recent -23.25 thousand episode would land below the interval.

Derived run: the pointwise median of the CDFs of 3 independent thesis.analyst fast rollouts on this target — median prediction sampling per Turtel et al. 2025 (arXiv:2505.17989). No new model call; this run is a deterministic aggregate of the recorded rollouts at 2026-07-10T17:19:37Z, 2026-07-10T17:25:21Z, 2026-07-10T17:30:53Z. Drivers and resolver fields mirror the rollout closest to the median.

Median CDF quantiles: q10 = 529.3, q50 = 541.5, q90 = 554.5. Constituent points [542.0, 541.5, 539.2] with 80% widths [24.8, 24.0, 33.2]; the median interval inherits the central rollout mass rather than averaging tails.

Framing: the resolver is Statistics Canada Table 14-10-0011-01, regular Employment Insurance beneficiaries, Canada, seasonally adjusted, beneficiary detail total, sex total, age group total, first print for June 2026. The target is in thousands; all anchors below are the same SA regular-beneficiary variant, not claims, unemployed persons, or unadjusted counts. I note a ledger text discrepancy: sourceBinding.table names the May 2026 release, but the dataPointId and resolutionRule correctly specify June 2026.

Reference class and base rate: for this exact series, the nearest official history is a four-month band from 542.11 to 554.83 thousand, with April at 544.44 thousand and March at 548.00 thousand. That base rate puts a no-shock June first print around the low-to-mid 540s rather than a return to the November 2025 peak of 569.00 thousand mentioned in the March release.

Prior/update/interval: persistence prior is April 2026 at 544.44 thousand, using the January-April 2026 official sample of 554.83, 542.11, 548.00, and 544.44 thousand; level effect holds near April, momentum effect is slightly negative after January and February declines and April's -0.5%, one-off release noise allows May/June catch-up, and policy-mechanism effect is neutral because temporary EI measures and eligibility/exhaustion mechanics can offset LFS improvement. The rung span is anchored by the fetched 542.11-548.00 thousand recent center, 554.83 thousand January upper recent print, and 569.00 thousand November 2025 peak, with downside room below 525 thousand if continuing exits dominate. Interval method is the elicited threshold ladder below, not a round symmetric band.

Ladder: P(X <= 515) = 0.03; P(X <= 520) = 0.07; P(X <= 525) = 0.12; P(X <= 530) = 0.20; P(X <= 535) = 0.32; P(X <= 540) = 0.46; P(X <= 545) = 0.60; P(X <= 550) = 0.72; P(X <= 555) = 0.82; P(X <= 560) = 0.89; P(X <= 565) = 0.94; P(X <= 570) = 0.97; P(X <= 575) = 0.99. Linear interpolation gives 10th percentile at 523.00, median at 541.43, and 90th percentile at 561.00, all in thousands.

Counter-considerations: upside risk is that May and June show renewed benefit entry from tariff-exposed goods jobs or delayed claims, which would land above the interval if the first print exceeds 561.00 thousand. Downside risk is faster exhaustion or return-to-work after the June LFS improvement, which would land below the interval if the first print is under 523.00 thousand. Outside the interval would require a two-month move larger than recent official month-to-month changes and not just ordinary first-print noise.

Review disposition: accepted the resolver wording fix by removing rounding as a resolution instruction, retained the May-versus-June sourceBinding.table discrepancy note for clarity, and made no change to the ladder-derived point or interval.

The target is the first Statistics Canada print for June 2026 in Table 14-10-0011-01: Canada, seasonally adjusted, regular benefits, beneficiary detail total, sex total, and age group total. Table values are persons and are converted to thousands by multiplying by 0.001. Later revisions are excluded.

The base rate is level persistence: before considering current labour-market or policy evidence, the prior is April's 544.44 thousand. The same-variant February–April first-print reference class is 542.11, 548.00, and 544.44 thousand, a 5.89-thousand range that anchors the dense central ladder rungs.

The registered August 20 resolution date is retained to keep the forecast tied to the canonical target despite the official schedule showing August 19. The sourceBinding description also says May 2026, metadata inconsistent with the June 2026 target; the dataPointId, resolver, and Table 14-10-0011-01 identify June unambiguously.

Prior/update/interval: The level-persistence model starts at 544.44 thousand, using the February–April first-print sample of 542.11, 548.00, and 544.44. The update adds about 6.5 thousand for lagged claims following the January–April employment decline and about 2.5 thousand for longer duration under temporary EI measures, then subtracts about 3.9 thousand for May's 88,000 employment gain and 6.6% unemployment, yielding an intended center near 549.5 thousand. The 10% lower tail allows rapid re-employment and benefit exhaustion to pull the level toward 527.5; the 10% upper tail allows delayed claims, longer durations, or concentrated losses to lift it toward 573.57. Direct threshold-ladder inversion gives final 80% bounds of 527.50 to 573.57 thousand.

Ladder: P(X <= 515) = 0.03; P(X <= 520) = 0.05; P(X <= 525) = 0.08; P(X <= 530) = 0.12; P(X <= 535) = 0.18; P(X <= 540) = 0.27; P(X <= 545) = 0.39; P(X <= 550) = 0.51; P(X <= 555) = 0.62; P(X <= 560) = 0.72; P(X <= 565) = 0.79; P(X <= 570) = 0.85; P(X <= 575) = 0.92; P(X <= 580) = 0.96. Linear interpolation gives the 10th percentile at 527.50, median at 549.58, and 90th percentile at 573.57.

Upside risk comes from delayed claims following early-2026 employment weakness, longer benefit duration under temporary EI measures, or concentrated manufacturing losses; an abrupt deterioration would land above the interval. Downside risk comes from sustained job gains, rapid claimant re-employment, or benefit exhaustion; a sharp normalization would land below the interval. A large administrative or policy-driven discontinuity is the principal outside the interval scenario.

Review disposition: Accepted the requested prior-first ordering, reconciled the unemployment driver to the cited May 6.6% figure, decomposed the center adjustment, allocated the interval tails to explicit scenarios, and flagged both ledger metadata discrepancies.

The resolver is Table 14-10-0011-01, series variant regular benefits, Canada, both sexes, age 15 years and over, seasonally adjusted; all anchors below use that same variant. The table is in persons and the target converts it to thousands. The ledger specifies 2026-08-20, while Statistics Canada's 2026 release-calendar PDF lists Employment Insurance for June on August 21; this concrete date discrepancy is retained against the supplied target contract rather than silently changing it.

Prior/update/interval: The persistence-with-recent-trend prior uses the fetched December-to-April reference class (567.62, 559.02, 550.35, 547.44, 544.44 thousand), whose decline decelerated into March-April. I update modestly lower for the May 6.6% unemployment rate, but retain substantial administrative and seasonal-adjustment uncertainty. The elicited ladder spans 520 to 580 thousand; its 10th and 90th interpolations set the 80% interval.

Ladder: P(X <= 520) = 0.03; P(X <= 525) = 0.06; P(X <= 530) = 0.11; P(X <= 535) = 0.20; P(X <= 540) = 0.39; P(X <= 545) = 0.59; P(X <= 550) = 0.68; P(X <= 555) = 0.80; P(X <= 560) = 0.88; P(X <= 565) = 0.93; P(X <= 570) = 0.96; P(X <= 575) = 0.98; P(X <= 580) = 0.99. Linear interpolation gives 10th percentile at 529.17, median at 542.75, and 90th percentile at 561.00 thousand.

upside risk: a renewed rise in job separations or weaker June hiring would lift beneficiaries above the central estimate. downside risk: continued re-employment and fewer new claims would reduce the level. Outside the interval: a broad labour-market deterioration would land above 561.00 thousand, while an unusually sharp benefit exit or claim decline would land below 529.17 thousand.

Review disposition: Accepted the blocking resolver-date critique. The official calendar supports August 21, while this cell remains bound to the supplied August 20 canonical target; the target must be corrected and re-registered before publication rather than changing its resolver fields independently.

The target is the seasonally adjusted Canada total for regular benefits, both sexes, age 15 years and over, in Statistics Canada Table 14-10-0011-01. Resolution is tied to the first official June 2026 print on 2026-08-20. The source-binding metadata is corrected here from May 2026 to June 2026 while retaining the canonical June target and first-print rule.

The base rate is persistence around the recent 2025-26 level, with the reference class consisting of the fetched June-through-October seasonal observations and the latest January-April sequence. The December-to-April decline was 23.18 thousand, while the January-April range was 8.73 thousand, so a return near the high-540s is more plausible than either a rapid collapse or a return to 570-plus.

The persistence prior is the primary model prior. I considered a simple time-series trend prior, but rejected extrapolating the December-to-April decline because the sequence includes a February low followed by a March rebound and April's smaller easing; a trend-only model would overstate downside. Level is anchored on April's 544.44 thousand, momentum is mixed, the one-off component is summer churn and expiry/re-entry timing, and no fetched source indicates a policy shock.

The May labour-market context is consistent with restraint on the upside: Statistics Canada's May 2026 indicators show employment at 21,122,000, up 0.4% monthly, and unemployment at 6.6%, down 0.3 percentage points. This supports a modest June rise rather than a sharp acceleration.

Prior/update/interval: use a persistence prior centered on the latest official April value of 544.44 thousand, update upward toward the fetched June 2025 value of 547.70 thousand and the 2025 summer values of 555.09, 555.27, and 554.27 thousand, and discount the improved May labour-market signal. The threshold-ladder interval method gives final implied bounds of 537.00 to 560.50 thousand and a median of 548.00 thousand.

Ladder: P(X <= 535) = 0.06; P(X <= 538) = 0.12; P(X <= 541) = 0.22; P(X <= 544) = 0.34; P(X <= 547) = 0.46; P(X <= 550) = 0.58; P(X <= 553) = 0.69; P(X <= 556) = 0.79; P(X <= 559) = 0.87; P(X <= 562) = 0.93; P(X <= 565) = 0.97. Linear interpolation gives 10th percentile at 537, median at 548, and 90th percentile at 560.5; values are in thousands and correspond to 537,000, 548,000, and 560,500 persons.

Downside risk is faster normalization after improved employment, which would land below the interval near 535 thousand. Upside risk is a stronger seasonal increase or delayed benefit exits, which would land above the interval near 565 thousand. A large unanticipated policy or labour-market shock would be outside the interval.

Review disposition: accepted the request for an explicit persistence-versus-time-series prior comparison, corrected the source-binding description to June 2026, and changed 'correction' to 'decline'; the ladder and forecast remain coherent with the fetched reference class.