# Employer Job Postings in Europe and North America: JOA Job Market Report, September 2026

*An analysis of the 1,385,899 open employer-direct job postings located in Europe and North America, from a snapshot of 2,111,453 postings by 144,098 employers: where the openings are, how often pay is stated, how old stated posting dates are and how concentrated postings are*

- Edition: September 2026
- Data as of: 2026-10-04
- Published: 2026-10-04 (updated 2026-10-04)
- Canonical page: https://jobopportunitiesapi.org/reports/job-market-september-2026
- Markdown version: https://jobopportunitiesapi.org/reports/job-market-september-2026.md
- Licence: [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)
- Cite as: Job Opportunities API (2026). JOA Job Market Report, September 2026 edition (snapshot of 4 October 2026). https://jobopportunitiesapi.org/reports/job-market-september-2026. CC BY 4.0.
- Aggregate data (CSV, 803 rows): https://jobopportunitiesapi.org/reports/job-market-september-2026/joa-job-market-2026-09.csv
- Data dictionary: https://jobopportunitiesapi.org/reports/job-market-september-2026/joa-job-market-2026-09-dictionary.md
- Audio narration (45 minutes 36 seconds, Kokoro bf_emma (synthetic)): https://jobopportunitiesapi.org/reports/job-market-september-2026#listen

## Summary

On 4 October 2026, JOA tracked 1,385,899 open employer-direct job postings located in North America (986,641) and Europe (399,258). Pay transparency divides neighbouring markets: Austrian postings are about four times as likely as German ones to state a pay figure (4.4 to 1 on a blind, AI-assisted sampled estimate, interval 2.9 to 6.2; 6.9 to 1 on automated detection), and in the United States 24.4% of own-site postings in the 10 jurisdictions with a pay-range posting law carry pay in a structured field, against 18.7% in 28 states without one. Stated posting dates are oldest in the Netherlands, where 43.9% of dated postings carry a stated posting date more than 90 days old, against 18.7% in Canada. Across the 51 US jurisdictions, JOA's postings follow the official JOLTS distribution by state (rank correlation 0.96); across 14 countries they are equivalent to 6.2% to 19.9% of official vacancy counts.

## Key facts

- **1,385,899** Open postings in Europe and North America (North America 986,641, Europe 399,258; 2,111,453 in total, 4 October 2026)
- **4.4 to 1** Pay stated: Austria vs Germany (sampled estimate, interval 2.9 to 6.2; 6.9 to 1 detected)
- **24.4% vs 18.7%** US own-site postings with pay (pay-range law jurisdictions vs other states)
- **6.2% to 19.9%** JOA postings as a share of official vacancies (14 countries; US states rank correlation 0.96 with JOLTS)

## 1. Key findings

- **Where the openings are.** North America accounts for 986,641 open postings (the United States 950,876, Canada 35,765) and Europe for 399,258, led by Germany (117,969), the United Kingdom (62,691), France (44,428) and the Netherlands (41,565).
- **Pay transparency in Europe.** Austrian postings are about four times as likely as German ones to state a pay figure: 4.4 to 1 on a blind, AI-assisted sampled estimate (interval 2.9 to 6.2; 78% against 18%) and 6.9 to 1 on automated detection (33.7% against 4.9%, detected by JOA's automated reading; the sampled estimate of postings that state pay is higher). Austria has required job adverts to state a minimum pay since 2011; Germany has no such rule.
- **Pay in structured fields on employers' careers pages** is most common in the United States (18.7%), the Netherlands (18.5%) and Canada (18.2%); Austria reaches 36.2% once pay written in the German text is counted.
- **US pay-range laws.** In the 10 jurisdictions with a pay-range posting law, 24.4% of own-site postings carry pay in a structured field, against 18.7% in 28 states without one: an association, and states differ in other ways.
- **Stated posting dates.** The share of dated postings whose stated posting date is more than 90 days old is highest in the Netherlands (43.9%) and Switzerland (38.0%) and lowest in Canada (18.7%); each recruiting system defines that date, and the share ranges from 2.2% on Workday to 65.6% on Personio.
- **Concentration.** The ten employers with the most open postings hold 71.9% of the 6,115 postings JOA tracks in Hungary, against 6.0% in Sweden and 8.9% in the United States.
- **Work arrangement.** Among postings on Lever, Ashby and Join that state a work model, hybrid leads in France (55%) and the United Kingdom (50%); on SmartRecruiters, used heavily by large employers, on-site dominates (81% in the United States).

**About the data.** This report analyses the job postings that employers publish on their own recruiting systems and careers pages (and some public employment service feeds, 33% of Sweden's postings), as available through JOA's API (jobopportunitiesapi.org) at 04:03 UTC on 4 October 2026: 2,111,453 open postings from 144,098 employers, of which 1,385,899 are located in Europe and North America and 560,385 have no resolved location and count only in totals. It is published by JOA, a one-person data business that sells API access ([About this report](https://jobopportunitiesapi.org/reports/job-market-september-2026#about)). Definitions, coverage and checks are in the [Methodology](https://jobopportunitiesapi.org/reports/job-market-september-2026#method); how this source differs from job boards, aggregators and official statistics is set out [there too](https://jobopportunitiesapi.org/reports/job-market-september-2026#m-sources).

## 2. Europe: overview

On 4 October 2026 JOA tracked 399,258 open postings in Europe from 49,597 employers. Germany is the largest market in the data with 117,969 postings (30% of the European total), followed by the United Kingdom (62,691), France (44,428) and the Netherlands (41,565). Employers in France post the most roles each (13.1 open postings per employer), against 5.7 in Germany. 56% of European postings come from employers' own careers pages and the rest from applicant-tracking systems; 67% carry a posting date, against 83% in North America.

**Open postings by region, 4 October 2026**

| Region | Open postings | Employers | Own careers pages | Has posting date | Date > 90 days old | Pay in structured field | Work model stated | Top-10 employers share |
| :--- | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: |
| Europe | 399,258 | 49,597 | 56% | 67% | 27.6% | 4.3% | 14.6% | 6.5% |
| North America | 986,641 | 79,355 | 58% | 83% | 21.2% | 11.7% | 9.0% | 8.6% |
| Other countries | 165,169 | 11,999 | 28% | 89% | 22.8% | 0.9% | 22.2% | 11.7% |
| Country not resolved | 560,385 | 30,947 | 43% | 77% | 17.8% | 1.0% | 8.6% | 6.1% |

Europe = EU27, EFTA (Switzerland, Norway, Iceland, Liechtenstein) and the United Kingdom. North America = United States and Canada (JOA's website also counts Mexico in its North America totals). 'Date > 90 days old' is a share of postings that carry a posting date. Employers are distinct company records within each row.

Compared with North America, European postings are older on average (27.6% of dated postings carry a date more than 90 days old, against 21.2%), state pay in a structured field less often (4.3% against 11.7%) and state a work model more often (14.6% against 9.0%). The ten employers with the most postings hold 6.5% of Europe's total. Country detail follows in section 4.

### What kinds of roles are open in Europe

Engineering is the largest role family in Europe (13.5% of classified postings), followed by Operations & Admin (11.2%) and Skilled Technician (10.3%).

**Largest role families: Europe and its largest markets**

| Region or country | Open postings | Largest | Second | Third | Unclassified |
| :--- | ---: | :--- | :--- | :--- | ---: |
| Europe | 399,258 | Engineering (13.5%) | Operations & Admin (11.2%) | Skilled Technician (10.3%) | 8.2% |
| Germany | 117,969 | Skilled Technician (14.4%) | Healthcare (10.8%) | Engineering (10.5%) | 10.2% |
| United Kingdom | 62,691 | Operations & Admin (15.4%) | Engineering (14.2%) | Sales (9.7%) | 7.9% |
| France | 44,428 | Engineering (15.6%) | Operations & Admin (12.5%) | Skilled Technician (9.4%) | 6.0% |
| Netherlands | 41,565 | Skilled Technician (14.4%) | Operations & Admin (13.6%) | Healthcare (11.2%) | 8.6% |
| Spain | 15,719 | Engineering (20.0%) | Sales (11.8%) | Operations & Admin (8.5%) | 6.0% |
| Italy | 9,416 | Sales (17.5%) | Engineering (12.4%) | Operations & Admin (12.0%) | 7.6% |
| Austria | 7,666 | Skilled Technician (13.3%) | Hospitality (11.6%) | Sales (10.3%) | 8.6% |

Shares of postings with a role family (classified from the job title at confidence 0.6 or higher; section 7). The mix reflects the employers that publish through the recruiting systems and careers pages JOA reads.

### Benchmark: JOA against official vacancy statistics

Official job vacancy statistics are the reference for vacancy levels, and JOA's counts are compared with them as a consistency check. They count vacancies at all employers, from employer surveys [1, 2, 3]; JOA counts the postings employers publish on their own recruiting systems and careers pages. Across the 14 countries with a recent official count, the rank correlation between JOA's open postings and official vacancies is 0.93 (Spearman; 0.97 for the logarithms), mostly reflecting country size. In level, JOA's open postings are equivalent to 6.2% to 19.9% of the official counts. A gap of this kind is expected: official counts include posts filled through agencies, personal networks and public employment services and posts never advertised online [4, 5], and the share of postings with a resolved country differs by recruiting system. Section 3 compares US states and Canadian provinces; section 7 sets out the methodological differences.

![Bar chart of JOA open postings divided by the latest official vacancy count, by country.](https://jobopportunitiesapi.org/reports/job-market-september-2026/charts/fig-coverage-index.svg)

**JOA open postings as a share of official vacancy counts** n = JOA open postings per country; official counts as listed in the table Source: JOA Job Market Report, September 2026; data as of 4 October 2026; CC BY 4.0; jobopportunitiesapi.org/reports

Ratios: Ireland 19.9%, Sweden 19.8%, United States 13.4%, Germany 11.5%, Netherlands 10.7%, Spain 10.4%, Switzerland 10.3%, France 9.2%, Belgium 9.1%, United Kingdom 8.9%, Poland 7.9%, Greece 7.9%, Canada 7.1%, Austria 6.2%. Official periods and measures differ between countries.

**JOA open postings and the latest official vacancy count, by country**

| Country | JOA open postings | Official vacancies | Official period | JOA as a share of official | Official measure and source |
| :--- | ---: | ---: | :--- | ---: | :--- |
| Ireland | 5,981 | 30,000 | Q2 2026 (prov.) | 19.9% | Eurostat job vacancies, provisional [1] |
| Sweden | 24,160 | 122,200 | Q2 2026 | 19.8% | Eurostat job vacancies [1] |
| United States | 950,876 | 7,079,000 | Aug 2026 (prelim.) | 13.4% | job openings, seasonally adjusted [6] |
| Germany | 117,969 | 1,023,400 | Q2 2026 | 11.5% | IAB job vacancy survey [2] |
| Netherlands | 41,565 | 387,200 | Q2 2026 | 10.7% | Eurostat job vacancies [1] |
| Spain | 15,719 | 151,736 | Q2 2026 | 10.4% | INE vacancies [7] |
| Switzerland | 8,840 | 86,034 | Q4 2025 | 10.3% | Eurostat job vacancies (latest found) [1] |
| France | 44,428 | 484,441 | Q2 2026 | 9.2% | all firm sizes, not seasonally adjusted [8] |
| Belgium | 12,133 | 133,881 | Q2 2026 | 9.1% | Eurostat job vacancies [1] |
| United Kingdom | 62,691 | 702,000 | Jun-Aug 2026 | 8.9% | vacancies, 3-month average [3] |
| Poland | 7,799 | 98,400 | Q2 2026 | 7.9% | Eurostat job vacancies [1] |
| Greece | 3,670 | 46,368 | Q2 2026 | 7.9% | ELSTAT job vacancies [9] |
| Canada | 35,765 | 501,035 | Jul 2026 | 7.1% | job vacancies, seasonally adjusted [10] |
| Austria | 7,666 | 123,400 | Q2 2026 | 6.2% | Statistik Austria open positions [11] |

JOA: snapshot of 4 October 2026. Official figures are survey-based counts of vacancies at all employers, for the period shown. Sweden includes 7,981 postings from the Swedish public employment service feed; without them its figure is 13.2%. Italy publishes a vacancy rate only (1.5% in Q2 2026) [12].

## 3. North America: overview

North America accounts for 986,641 open postings from 79,355 employers, 96% of them in the United States (950,876 postings from 76,827 employers) and the rest in Canada (35,765 from 4,797). US employers average 12.4 open postings each, against 7.5 in Canada. 58% of North American postings come from employers' own careers pages, 83% carry a posting date, and 11.7% state pay in a structured field, more than twice the European share. The ten employers with the most postings hold 8.6% of the region's total.

### What kinds of roles are open in North America

Healthcare is the largest role family in North America (17.4% of classified postings), followed by Operations & Admin (11.4%) and Engineering (8.7%).

**Largest role families: North America, the United States and Canada**

| Region or country | Open postings | Largest | Second | Third | Unclassified |
| :--- | ---: | :--- | :--- | :--- | ---: |
| North America | 986,641 | Healthcare (17.4%) | Operations & Admin (11.4%) | Engineering (8.7%) | 7.2% |
| United States | 950,876 | Healthcare (17.8%) | Operations & Admin (11.4%) | Retail (8.6%) | 7.2% |
| Canada | 35,765 | Engineering (14.4%) | Operations & Admin (11.6%) | Sales (8.9%) | 7.6% |

Shares of postings with a role family (classified from the job title at confidence 0.6 or higher; section 7). The mix reflects the employers that publish through the recruiting systems and careers pages JOA reads.

### Benchmark: US states and Canadian provinces against official statistics

Nationally, JOA's US postings are equivalent to 13.4% of JOLTS job openings for Aug 2026 (prelim.) and its Canadian postings to 7.1% of Statistics Canada's vacancies for Jul 2026. The distribution across states is consistent with JOLTS: across the 51 jurisdictions with a JOLTS state estimate, the rank correlation between JOA's open postings and JOLTS openings is 0.96. The largest differences in share of the US total are Ohio (-1.1 points), Georgia (-1.1 points) and California (+1.0 points). Relative to its JOLTS share, a state's JOA share ranges from 0.5 times (West Virginia) to 1.7 times (New Hampshire), with a median of 1.00; the JOLTS state values are for December 2025.

**US states: the ten largest by JOLTS job openings, JOA and JOLTS shares of the US total**

| State | JOA open postings | JOA share | JOLTS openings (Dec 2025) | JOLTS share | Difference |
| :--- | ---: | ---: | ---: | ---: | ---: |
| California | 83,714 | 11.2% | 674,000 | 10.3% | +1.0 points |
| Texas | 72,244 | 9.7% | 608,000 | 9.3% | +0.4 points |
| New York | 46,642 | 6.3% | 431,000 | 6.6% | -0.3 points |
| Florida | 44,487 | 6.0% | 404,000 | 6.2% | -0.2 points |
| Ohio | 24,140 | 3.2% | 287,000 | 4.4% | -1.1 points |
| Georgia | 23,627 | 3.2% | 279,000 | 4.3% | -1.1 points |
| Pennsylvania | 29,220 | 3.9% | 235,000 | 3.6% | +0.3 points |
| Michigan | 18,691 | 2.5% | 221,000 | 3.4% | -0.9 points |
| Virginia | 23,317 | 3.1% | 213,000 | 3.3% | -0.1 points |
| North Carolina | 21,650 | 2.9% | 212,000 | 3.2% | -0.3 points |

Shares of the total across all 51 jurisdictions (50 states and the District of Columbia): 744,347 JOA open postings that carry a state, and 6,550,000 JOLTS job openings, December 2025, the latest state value published at the time of the snapshot [6]. The full table is in the CSV download.

In Canada, among the 16.7% of postings that carry a province, Ontario, Quebec, British Columbia and Alberta rank in the same order as in the official data, with Ontario's share higher in JOA (39.6% against 32.9%).

**Canadian provinces: shares of JOA postings with a province and of official vacancies**

| Province | JOA open postings | JOA share | Official vacancies (Jul 2026) | Official share |
| :--- | ---: | ---: | ---: | ---: |
| ON | 2,365 | 39.6% | 165,935 | 32.9% |
| QC | 1,236 | 20.7% | 115,960 | 23.0% |
| BC | 954 | 16.0% | 89,545 | 17.8% |
| AB | 799 | 13.4% | 70,860 | 14.1% |
| NS | 222 | 3.7% | 12,665 | 2.5% |
| MB | 130 | 2.2% | 20,660 | 4.1% |
| SK | 111 | 1.9% | 13,900 | 2.8% |
| NB | 122 | 2.0% | 10,185 | 2.0% |
| NL | 29 | 0.5% | 4,040 | 0.8% |

JOA shares are of the 16.7% of Canadian postings that carry a province [10].

## 4. Country and state deep dives

The table sets the fifteen countries in this edition's country set side by side; the profiles below pick out what distinguishes each market.

**Open postings by country, 4 October 2026**

| Country | Open postings | Employers | Postings per employer | Own careers pages | Has posting date | Date > 90 days old | Pay in structured field | Pay incl. German text | Top-10 employers share |
| :--- | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: |
| United States | 950,876 | 76,827 | 12.4 | 58% | 83% | 21.3% | 11.9% | 11.9% | 8.9% |
| Canada | 35,765 | 4,797 | 7.5 | 38% | 85% | 18.7% | 8.6% | 8.6% | 12.0% |
| Germany | 117,969 | 20,783 | 5.7 | 79% | 53% | 32.4% | 3.9% | 4.9% | 14.7% |
| United Kingdom | 62,691 | 9,323 | 6.7 | 49% | 74% | 20.2% | 5.4% | 5.4% | 9.8% |
| France | 44,428 | 3,398 | 13.1 | 41% | 80% | 24.6% | 5.5% | 5.5% | 16.7% |
| Netherlands | 41,565 | 6,487 | 6.4 | 62% | 52% | 43.9% | 11.5% | 11.5% | 8.7% |
| Spain | 15,719 | 2,398 | 6.6 | 35% | 88% | 26.6% | 2.5% | 2.5% | 12.7% |
| Italy | 9,416 | 1,360 | 6.9 | 36% | 80% | 19.8% | 2.1% | 2.1% | 22.4% |
| Austria | 7,666 | 1,633 | 4.7 | 61% | 62% | 24.8% | 7.1% | 33.7% | 23.8% |
| Greece (thin sample) | 3,670 | 571 | 6.4 | 19% | 83% | 33.2% | 0.9% | 0.9% | 23.2% |
| Sweden | 24,160 | 5,256 | 4.6 | 25% | 95% | 21.4% | 0.1% | 0.1% | 6.0% |
| Belgium | 12,133 | 2,045 | 5.9 | 65% | 50% | 33.3% | 0.9% | 0.9% | 28.2% |
| Switzerland | 8,840 | 1,600 | 5.5 | 65% | 72% | 38.0% | 0.5% | 0.5% | 25.8% |
| Poland | 7,799 | 1,474 | 5.3 | 28% | 86% | 22.0% | 2.4% | 2.4% | 14.1% |
| Ireland | 5,981 | 1,172 | 5.1 | 41% | 76% | 21.1% | 2.8% | 2.8% | 17.3% |

Pay figures are those detected by JOA's automated reading (section 5.2); 'Pay incl. German text' adds pay that JOA reads from German-language advert text (Austria and Germany). Percentages need a base of at least 100 postings. Greece: 3,670 postings and 33 with a pay figure, so its pay and work-model shares rest on small numbers. Within Europe, Hungary (6,115 open postings), Norway (5,747), Portugal (4,706), Denmark (3,538) and Finland (3,097) also have more than 3,000 open postings; Hungary and Norway appear in the pay and concentration charts and all are in the CSV.

### Country profiles

- **Germany** is Europe's largest market in the data: 117,969 open postings from 20,783 employers. 79% come from employers' own careers pages, the highest share in the country set. 53% carry a posting date, and 32.4% of those are more than 90 days old. JOA's automated reading detects a pay figure in 4.9% of postings (section 5).
- **The United Kingdom** has 62,691 postings from 9,323 employers, with one of the lowest shares of stated posting dates more than 90 days old in Europe (20.2%) and low concentration (top ten employers 9.8%). Hybrid is the most common stated work model on recruiting software popular with tech employers (Lever, Ashby, Join; 50%).
- **France** (44,428 postings) has the most postings per employer in Europe (13.1) and the highest hybrid share on that software (55%). 12.8% of own-site postings carry pay in a structured field.
- **The Netherlands** (41,565 postings) has the highest share of stated posting dates more than 90 days old: 43.9% of dated postings, on the 52% that carry a date. Its own-site postings state structured pay as often as those in the United States and Canada (18.5%).
- **Spain** (15,719) and **Italy** (9,416) have high date coverage (88% and 80%) and low pay disclosure (2.5% and 2.1%). In Spain on-site leads among postings on Lever, Ashby and Join that state a work model (56%); in Italy the ten largest employers hold 22.4% of postings.
- **Austria** (7,666 postings from 1,633 employers) stands out on pay: 33.7% of postings carry a detected pay figure, the highest share in Europe (section 5).
- **Sweden** (24,160) has the highest share of dated postings (95%) and the least concentrated postings (top ten 6.0%); 33% of Swedish postings come from the public employment service feed.
- **Switzerland** (8,840) and **Belgium** (12,133) combine older dates (38.0% and 33.3% more than 90 days old) with high concentration (top ten 25.8% and 28.2%). **Poland** (7,799) and **Ireland** (5,981) sit near the European middle on both.
- **Hungary**, outside the country table, has 6,115 postings, of which the ten largest employers hold 71.9%, the highest concentration in the data; Hungarian figures therefore reflect a small group of employers.
- **Canada** (35,765 postings from 4,797 employers) has the lowest share of stated posting dates more than 90 days old in the two regions (18.7%) and 18.2% structured pay on own-site postings. 16.7% of Canadian postings carry a province (section 3).

### US states

California (83,714 postings) and Texas (72,244) lead, followed by New York (46,642), Florida (44,487) and Illinois (29,320). The ten largest states hold 53.5% of US postings that carry a state, in line with their weight in the official JOLTS estimates (section 3). Structured pay on own-site postings ranges from 16.5% to 27.2% among them.

**US states with the most open postings**

| State | Open postings | Share of US postings with a state | Own-site postings | Own-site postings with structured pay | Pay-law group (section 5) |
| :--- | ---: | ---: | ---: | ---: | :--- |
| California | 83,714 | 11.2% | 45,562 | 25.5% | Pay-range posting law |
| Texas | 72,244 | 9.7% | 43,187 | 16.5% | No posting-range law found |
| New York | 46,642 | 6.3% | 22,218 | 25.2% | Pay-range posting law |
| Florida | 44,487 | 6.0% | 27,926 | 20.1% | No posting-range law found |
| Illinois | 29,320 | 3.9% | 16,598 | 19.5% | Pay-range posting law |
| Pennsylvania | 29,220 | 3.9% | 16,344 | 27.2% | No posting-range law found |
| Ohio | 24,140 | 3.2% | 14,578 | 16.9% | Law unconfirmed or on request only |
| Georgia | 23,627 | 3.2% | 13,886 | 20.1% | No posting-range law found |
| Virginia | 23,317 | 3.1% | 12,623 | 22.3% | Law unconfirmed or on request only |
| North Carolina | 21,650 | 2.9% | 12,094 | 21.2% | No posting-range law found |

Shares of the 744,347 US postings that carry a state (205,370 US postings carry no state). All states are in the CSV.

## 5. Pay transparency: the law and the postings

Pay transparency is the clearest dividing line in this edition. This section sets out what the law requires on each side of the Atlantic, then what employer-direct postings show.

### 5.1 What the law requires

The EU Pay Transparency Directive (2023/970) gives job applicants a right to pay information before the interview and names a published job advert as one way to provide it, so an advert without pay can comply in countries that follow the Directive's wording [13]. Member states had to transpose it by 7 June 2026. Reporting duties on pay gaps start in 2027 for employers with 150 or more workers and in 2031 for those with 100 to 149.

As of 3 October 2026, the EUR-Lex page of national measures communicated to the Commission lists none for Germany, France, the Netherlands, Ireland, Denmark, Croatia, Cyprus, Latvia, Luxembourg, Hungary, Portugal and Finland; Italy and Malta show one each, and Austria, Belgium, Greece, Poland, Spain and Sweden between 5 and 33 [14]. These counts record notification, and this report characterises no country as fully transposed or in breach; Austria's count, for instance, sits alongside a rule that predates the Directive.

**Europe: rules on pay information for job applicants**

| Jurisdiction | What the law says | In force | Source |
| :--- | :--- | :--- | :--- |
| European Union | Employers must give applicants the initial pay or its range, based on objective gender-neutral criteria, 'such as in a published job vacancy notice, prior to the job interview or otherwise'. The advert is one permitted channel. Pay history may not be asked. Notices and job titles must be gender-neutral. | Transposition due 7 June 2026 | [13] |
| Austria | An advert must state the minimum pay set by collective agreement or law, and whether the employer is willing to pay more. After a first warning, a fine of up to EUR 360. | 1 March 2011 (fine from 1 January 2012) | [15] |
| Poland | Applicants must be told the starting pay or range in the advert or before the interview (Labour Code art. 18(3ca)); partial transposition of the Directive. | 24 December 2025 | [16] |
| Germany | The Pay Transparency Act of 2017 gives an information right to employees and sets no duty to state pay in adverts. The official consolidated text shows no transposition act. | 2017 | [17] |
| Spain | Royal Decree 902/2020 on pay equality has no rule on pay in adverts; the 2026 transposition status is unconfirmed in official sources. | 2020 | [18] |
| Great Britain | Equality Act 2010 section 77 covers pay discussions; the sections checked (ss. 77-78) contain no duty to state pay in adverts. | - | [19] |
| Netherlands | Bill 36949 to implement the Directive is pending. | Pending | [20] |

Sources opened on 3 October 2026. The 2026 status of France, Italy, Belgium, Greece and Ireland, and the French Labour Code rule on pay information before the interview, are omitted pending confirmation from an official source. Factual context, not legal advice.

In North America, a growing number of US states and two Canadian provinces, British Columbia and Ontario, require employers to put a pay range in a job posting. Thresholds, exemptions and content differ; some laws also require benefits information.

**North America: jurisdictions that require a pay range in postings**

| Jurisdiction | Who is covered | In force | Source |
| :--- | :--- | :--- | :--- |
| California | 15+ employees | 1 January 2023 | [21] |
| Illinois | 15+ employees; pay and benefits | 1 January 2025 | [22] |
| Maryland | any employer | 1 October 2024 | [23] |
| Massachusetts | 25+ employees | 29 October 2025 (derived: one year after the act) | [24] |
| Minnesota | 30+ employees | 1 January 2025 | [25] |
| New Jersey | 10+ employees | 1 June 2025 | [26] |
| New York | 4+ employees | 17 September 2023 | [27] |
| Vermont | 5+ employees | 1 July 2025 | [28] |
| Washington | 15+ employees; pay and benefits | 1 January 2023 | [29] |
| District of Columbia | any employer | 30 June 2024 | [30] |
| British Columbia | no size threshold | 1 November 2023 | [31] |
| Ontario | 25+ employees; pay above CAD 200,000 exempt | 1 January 2026 (date read from the consolidated Act) | [32] |

Delaware has enacted such a law, effective 26 September 2027 [33]. Connecticut and Rhode Island require a range on request or at the offer stage [34]. Colorado is excluded from the comparison in this section until its statute text is verified; the legislature's bill summary describes a posting rule [35]. Newfoundland and Labrador's posting provision awaits proclamation.

Where pay must appear in the advert itself (Austria's minimum pay, the US jurisdictions and the two provinces above), a high share partly reflects compliance. Poland's rule and the Directive can also be met before the interview, and size thresholds mean a country or state share mixes covered and uncovered employers. The figures below measure disclosure in postings, not compliance with the law.

### 5.2 What employer-direct postings show: Austria and Germany

JOA detects a pay figure in 33.7% of the 7,666 open postings in Austria and 4.9% of the 117,969 in Germany (detected by JOA's automated reading; the sampled estimate of postings that state pay is higher, see below). The gap holds within every recruiting system with at least 150 Austrian postings, so it is not explained by which recruiting systems employers use.

![Grouped bar chart comparing Austria and Germany: share of open postings with a detected pay figure, overall and within each recruiting system.](https://jobopportunitiesapi.org/reports/job-market-september-2026/charts/fig-at-de-pay.svg)

**Postings with a pay figure: Austria and Germany** n: Austria 7,666, Germany 117,969 open postings; system rows have at least 150 Austrian postings Source: JOA Job Market Report, September 2026; data as of 4 October 2026; CC BY 4.0; jobopportunitiesapi.org/reports

Austria is far above Germany overall (33.7% against 4.9%) and in every recruiting system shown: Own careers pages 36.2% against 5.8%; SmartRecruiters 33.1% against 0.3%; Recruitee 28.4% against 1.9%; Workday 41.9% against 0.6%; Teamtailor 21.9% against 0.7%.

A blind, AI-assisted re-reading of stratified random samples (84 Austrian and 100 German postings) puts the share of postings that state a pay figure at 78% in Austria (95% interval 70% to 86%) and 18% in Germany (13% to 26%). Austrian adverts are about four times as likely as German ones to state a pay figure on the sampled estimate (4.4 to 1; interval 2.9 to 6.2), 6.9 to 1 on automated detection. Section 7 describes the sample.

**Austria and Germany: postings with a pay figure detected**

| Measure | Austria | Germany |
| :--- | ---: | ---: |
| Open postings | 7,666 | 117,969 |
| Pay in a structured field of the source | 7.1% (545) | 3.9% (4,639) |
| Pay read by JOA from German advert text | 2,039 | 1,091 |
| Any pay figure detected | 33.7% (2,584) | 4.9% (5,730) |
| States a pay figure, sampled estimate (95% interval) | 78% (70 to 86%) | 18% (13 to 26%) |

Counts of postings in brackets. A pay figure is positive and at most 1,000,000 euro a year after conversion. Sampled estimate: 84 Austrian and 100 German postings read independently (section 7).

The difference is consistent with the law. Austria has required adverts to state the collective-agreement or statutory minimum pay since 1 March 2011, with a fine of up to EUR 360 after a first warning [15]; Germany's pay-transparency law gives a right to information to employees and sets no duty for adverts [17]. Other differences between the two labour markets can contribute, and the pay wording JOA detects includes Austrian legal terms (collective agreement, minimum pay). An Austrian figure is often a floor: about 38% of the amounts JOA read from Austrian advert text are the collective-agreement minimum rather than the pay on offer. This report uses the presence of a pay figure and publishes no pay levels.

### 5.3 Pay in structured fields on employers' careers pages

To compare countries on a like-for-like basis, this part uses one source type: postings read from employers' own careers pages, where pay is often published as structured data. Countries need at least 2,500 such postings.

![Bar chart ranking countries by the share of open own-careers-page postings that carry pay in a structured field.](https://jobopportunitiesapi.org/reports/job-market-september-2026/charts/fig-pay-league.svg)

**Own-careers-page postings with pay in a structured field** n = open postings read from own careers pages (all countries with at least 2,500); \* share including pay read from German text shown in the label Source: JOA Job Market Report, September 2026; data as of 4 October 2026; CC BY 4.0; jobopportunitiesapi.org/reports

Ranking by structured pay: United States 18.7%, Netherlands 18.5%, Canada 18.2%, France 12.8%, Austria 11.0%, United Kingdom 10.1%, Spain 6.6%, Germany 4.9%, Italy 4.9%, Australia 3.7%, Belgium 1.3%, India 0.9%, Switzerland 0.8%, Norway 0.7%, Sweden 0.3%, Hungary 0.1%. Including pay read from German text, Austria reaches 36.2% and Germany 5.8%.

On structured pay, the United States (18.7%), the Netherlands (18.5%) and Canada (18.2%) lead, followed by France (12.8%), Austria (11.0%) and the United Kingdom (10.1%); Germany stands at 4.9%. Counting pay written in German advert text, Austria reaches 36.2% and Germany 5.8%. Belgium, Switzerland, Norway, Sweden and Hungary are below 3%; in Hungary ten employers hold 72% of postings, so its figure reflects a handful of employers. The Netherlands, which has no posting rule, is level with the United States and Canada, so law is one driver among several.

The Indeed Hiring Lab publishes the share of Indeed postings that include pay [36], and its August 2026 shares are higher than JOA's in every country compared. The two measure different things: Indeed counts pay information anywhere in its postings, while JOA's league counts pay in structured fields on employers' own pages only, so it understates how often the postings JOA tracks state pay. Research on US online postings before the recent state laws found wage information in about 14% of postings [37]; the share rises sharply once posting laws take effect [38].

![Grouped bar chart comparing the JOA own-careers-page structured pay share with the Indeed Hiring Lab share of postings with pay, for five countries.](https://jobopportunitiesapi.org/reports/job-market-september-2026/charts/fig-pay-vs-indeed.svg)

**Two ways of measuring pay in postings** JOA: own-careers-page postings, structured pay only, 4 October 2026 (Germany 4.9% here; 5.8% in the league chart including German text); Indeed: August 2026, rounded (see references) Source: JOA data: JOA Job Market Report, September 2026 (CC BY 4.0). Indeed data: Indeed Hiring Lab Pay Transparency Tracker, August 2026 (CC BY 4.0), rounded.

Indeed shares are higher in every country: United States JOA 18.7% against Indeed 51.8%; Canada JOA 18.2% against Indeed 57.6%; United Kingdom JOA 10.1% against Indeed 59.5%; Germany JOA 4.9% against Indeed 14.7%; France JOA 12.8% against Indeed 45.3%.

### 5.4 US states with pay-range posting laws

The comparison covers own-careers-page postings in the 10 jurisdictions with a pay-range posting law in force on 1 September 2026 (California, Illinois, Maryland, Massachusetts, Minnesota, New Jersey, New York, Vermont, Washington and the District of Columbia) and 28 states with no such law found. 13 states where the law is unconfirmed or requires a range only on request (Alaska, Arkansas, Colorado, Connecticut, Hawaii, Maine, Michigan, Nebraska, Nevada, Ohio, Oregon, Rhode Island and Virginia) sit outside both groups.

**Own-careers-page postings with pay in a structured field, US state groups**

| Group | States | Own-site postings | With pay | Share |
| :--- | ---: | ---: | ---: | ---: |
| Pay-range posting law in force | 10 | 135,939 | 33,108 | 24.4% |
| No posting-range law found | 28 | 219,603 | 40,995 | 18.7% |
| Outside both groups (law unconfirmed or range on request only) | 13 | 70,879 | 14,259 | 20.1% |

US postings without a state (205,370) are outside all three groups. Delaware is a comparison state: its law takes effect in 2027.

![Bar chart of the share of own-careers-page postings carrying structured pay, by US state, with law states highlighted.](https://jobopportunitiesapi.org/reports/job-market-september-2026/charts/fig-us-states.svg)

**US states: own-careers-page postings with pay in a structured field** n = open own-careers-page postings per state (states with at least 4,000 such postings) Source: JOA Job Market Report, September 2026; data as of 4 October 2026; CC BY 4.0; jobopportunitiesapi.org/reports

Law states: MA 26.6%, MD 25.8%, CA 25.5%, MN 25.4%, NY 25.2%, NJ 24.8%, WA 20.6%, IL 19.5%. Comparison states: PA 27.2%, MO 22.9%, KY 21.4%, NC 21.2%, SC 20.6%, FL 20.1%, GA 20.1%, AL 19.0%, UT 18.8%, IN 17.7%, OK 17.5%, WI 16.7%, TX 16.5%, KS 16.2%, TN 15.3%, LA 15.1%, AZ 13.8%.

The share is 24.4% in the law jurisdictions against 18.7% in the comparison states, a difference of 5.7 percentage points: an association, and states differ in other ways (national employers often publish one posting for all states, and laws differ in thresholds and exemptions). The pattern varies: Pennsylvania, with no law found, reaches 27.2%, while Illinois, with a law, stands at 19.5%. Research using other data finds that state posting laws raised the share of postings with pay by about 30 percentage points [38]; JOA's cross-sectional gap is a different quantity. The structured-field measure leaves out pay written only in the posting text, so it understates how often pay is stated.

Canada's provinces are compared in section 3 rather than here, because 16.7% of Canadian postings carry a province [31, 32].

## 6. Hiring dynamics

### 6.1 Age of stated posting dates

For postings with an employer-stated posting date, the chart shows the share whose stated posting date is more than 90 days old. Each recruiting system defines that date (first publication, last update or start date), so the figure describes the stated date as each system records it.

![Bar chart ranking countries by the share of dated open postings whose posting date is older than 90 days.](https://jobopportunitiesapi.org/reports/job-market-september-2026/charts/fig-age-90.svg)

**Open postings whose stated posting date is more than 90 days old** n = open postings with a posting date (countries in the country set, section 4, plus Australia as a reference, with at least 1,000) Source: JOA Job Market Report, September 2026; data as of 4 October 2026; CC BY 4.0; jobopportunitiesapi.org/reports

Ranking: Netherlands 43.9%, Switzerland 38.0%, Belgium 33.3%, Greece 33.2%, Germany 32.4%, Spain 26.6%, Austria 24.8%, France 24.6%, Poland 22.0%, Sweden 21.4%, United States 21.3%, Ireland 21.1%, United Kingdom 20.2%, Italy 19.8%, Canada 18.7%, Australia 13.6%.

The Netherlands (43.9%), Switzerland (38.0%) and Belgium (33.3%) have the highest shares; among European and North American countries Canada has the lowest (18.7%), and Australia, shown as a reference, 13.6%. On own-careers-page postings alone, the Netherlands and Switzerland stay highest. Recruiting systems define the posting date differently, and the share older than 90 days ranges from 2.2% on Workday to 65.6% on Personio, so part of a country's figure reflects its mix of systems. An old stated date can mark an evergreen posting, a re-used template or a long search.

### 6.2 Concentration among employers

Ranking employers within each country by their number of open postings, the chart shows the share held by the ten largest. No employer is named in this report.

![Bar chart ranking countries by the share of open postings held by their ten largest employers.](https://jobopportunitiesapi.org/reports/job-market-september-2026/charts/fig-concentration.svg)

**Share of open postings held by the ten largest employers** n = open postings per country (all countries in Europe and North America with at least 5,000, plus Australia and India as references) Source: JOA Job Market Report, September 2026; data as of 4 October 2026; CC BY 4.0; jobopportunitiesapi.org/reports

Ranking: Hungary 71.9%, India 29.6%, Norway 29.2%, Belgium 28.2%, Switzerland 25.8%, Austria 23.8%, Italy 22.4%, Ireland 17.3%, France 16.7%, Australia 15.1%, Germany 14.7%, Poland 14.1%, Spain 12.7%, Canada 12.0%, United Kingdom 9.8%, United States 8.9%, Netherlands 8.7%, Sweden 6.0%.

The ten largest employers hold 71.9% of the 6,115 open postings JOA tracks in Hungary and 29.2% in Norway, against 6.0% in Sweden; across whole regions they hold 6.5% in Europe and 8.6% in North America. Where a few employers hold a quarter or more of the postings, a country's averages move with those employers' hiring.

### 6.3 Work arrangement

15% of European and 9% of North American postings state a work model, and the share that states one differs widely between recruiting systems. Countries are therefore compared within one group of systems, Lever, Ashby and Join, recruiting software popular with tech employers, in which most postings state a work model.

![Stacked bar chart of remote, hybrid and on-site shares among postings that state a work model, by country.](https://jobopportunitiesapi.org/reports/job-market-september-2026/charts/fig-work-model.svg)

**Stated work model in postings on Lever, Ashby and Join** n = open postings that state a work model (countries in the country set, section 4, with at least 1,000) Source: JOA Job Market Report, September 2026; data as of 4 October 2026; CC BY 4.0; jobopportunitiesapi.org/reports

United States: remote 30%, hybrid 24%, on-site 46%; United Kingdom: remote 25%, hybrid 50%, on-site 26%; Canada: remote 30%, hybrid 36%, on-site 34%; Germany: remote 24%, hybrid 37%, on-site 39%; Spain: remote 19%, hybrid 25%, on-site 56%; France: remote 15%, hybrid 55%, on-site 30%.

Hybrid is the most common stated model, by at least three points, in the United Kingdom (50%) and France (55%). On-site leads in the United States (46%) and Spain (56%). The remaining countries are split closely between two models. Countries with fewer than 1,000 postings stating a work model (Australia, n=681, Netherlands, n=618, Poland, n=546) are in the CSV.

Among the large employers that use SmartRecruiters, on-site is by far the most common stated model:

**Stated work model on SmartRecruiters, used heavily by large employers**

| Country | Postings stating a work model | Remote | Hybrid | On-site |
| :--- | ---: | ---: | ---: | ---: |
| United States | 19,677 | 8% | 11% | 81% |
| France | 8,800 | 3% | 15% | 82% |
| Germany | 5,228 | 5% | 22% | 73% |
| United Kingdom | 4,928 | 5% | 22% | 73% |
| Canada | 3,459 | 5% | 21% | 75% |
| Netherlands | 1,515 | 3% | 20% | 76% |
| Austria | 1,428 | 1% | 8% | 91% |
| Spain | 1,395 | 9% | 29% | 62% |

A keyword-based tracker covering all postings on one large job site finds remote or hybrid wording in roughly 8% to 18% of postings in the countries it covers (August 2026) [39]; studies of posting text show remote-work advertising rising strongly since 2019 [40, 41]. JOA counts the model the employer states, on a different set of employers.

### 6.4 Weekly postings in September

A fixed panel of 1,739 European and 3,147 North American employers shows the pace of new postings by employer-stated posting date. The weekly series starts with the week of 31 August 2026, the first full week in which posting dates are captured consistently across the seven systems.

![Line chart of postings per week in Europe and North America by a fixed panel of employers, from the week of 31 August.](https://jobopportunitiesapi.org/reports/job-market-september-2026/charts/fig-weekly-panel.svg)

**Postings per week by a fixed panel of employers** Weeks: 31 Aug, 7 Sep, 14 Sep, 21 Sep. Panel employers: Europe 1,739, North America 3,147 Source: JOA Job Market Report, September 2026; data as of 4 October 2026; CC BY 4.0; jobopportunitiesapi.org/reports

Europe: 7,123, 7,077, 7,142, 7,654 per week. North America: 13,921, 10,635, 13,131, 13,305 per week, with a dip in the week of 7 September, which contains US Labor Day.

In North America the panel posted 13,921 roles in the week of 31 Aug and 13,305 in the week of 21 Sep (-4.4%), with a dip in the week of 7 September, which contains US Labor Day. In Europe it moved from 7,123 to 7,654 (+7.5%), all of the rise in the last week. Neither movement reaches the 10% this report requires before reading a change over four weeks as a trend.

### 6.5 Posting days

Of 495,241 postings with a posting date in complete weeks since 31 August, 4.2% are dated on a Saturday or Sunday, so about 96% are dated Monday to Friday. Posting dates are date-only and defined by each recruiting system.

## 7. Methodology and limitations

### Data and population

All figures come from one consistent read-only snapshot of the JOA database taken at 2026-10-04 04:03 UTC on 4 October 2026, and reflect that snapshot after automated and sampled quality checks. The population is the set of postings JOA's API serves to its customers: postings found live at an employer's own recruiting system or careers page (plus a small number from public employment service feeds), not opted out by the employer and passing JOA's quality checks. Postings collected more than once (same employer, title, city and apply link) and re-posts whose title begins 'Copy of' are held back from it, as are records from 19 other sources that JOA reads only to discover new postings (job boards, aggregators and some public employment service portals; listed on jobopportunitiesapi.org/coverage). 2,111,453 open postings from 144,098 employer records were analysed. JOA's public statistics (api.jobopportunitiesapi.org/public/site-stats, as of 4 October 2026, measured 2026-10-04 04:53 UTC) use the same definition and show 2,113,135 live postings from 144,132 employers; the report's counts differ slightly because its snapshot was taken at a different minute.

### How JOA's data differs from other sources

Three kinds of source describe job openings. They measure different things, and the differences explain most of what separates their numbers.

**Job boards and aggregators.** Job boards and aggregators, and the analytics built on them, collect postings from employers' own sites, other boards and recruitment agencies. Employers often publish the same vacancy on several portals and repost it while it is open, and an intermediary can advertise a role alongside the employer, so the same role can appear several times [42, 43, 44]. Counts therefore depend on de-duplication rules the reader cannot inspect: providers match postings on title, company, location or text within a time window and revise those rules over time [45, 46, 4, 47]. When an intermediary places the advert, the text often omits which company is hiring, so the industry is unknown [43]; on one technical-jobs board, recruitment and staffing firms accounted for two-thirds of all postings [48]. Filled or withdrawn positions can stay online until the site removes them, most commonly after 30 days, and an aggregator treats a posting as live while any board still shows it [49, 4, 47]. One advert can cover several vacancies [49, 47], and some employers advertise permanently or to build reserve lists of applicants [42]. Statistics Netherlands found that a de-duplicated count of scraped vacancies still differed too much from its vacancy survey to replace it [43].

**Employer-direct data (JOA).** JOA reads each role at the employer's own careers site or applicant-tracking system, with the employer's apply link, holds back copies collected more than once, re-checks postings daily and records a posting as closed when it leaves the employer's system. The employer is the one that publishes the posting; location comes from the employer's posting data (its accuracy is given under Location, below). Employer-direct data has its own limits: roles advertised only through agencies or job boards, employers on recruiting systems JOA does not yet read, and small employers without an applicant-tracking system or careers page are outside it; 26.5% of postings have no resolved location; and one posting can cover several vacancies or stay up for recurring hiring. Advert-based data in general over-represent higher-skilled and white-collar occupations and some industries [50, 51, 52, 53].

**Official statistics.** JOA's counts are compared with official job vacancy statistics, the reference for vacancy levels: Eurostat job vacancy statistics [1, 5], the ONS Vacancy Survey (United Kingdom) [3, 54], the IAB Job Vacancy Survey (Germany) [2, 55, 56], the DARES/INSEE Acemo survey (France) [8, 57], INE (Spain) [7, 58], ISTAT (Italy, vacancy rate) [12, 59], Statistik Austria [11, 60], ELSTAT (Greece) [9], BLS JOLTS (United States) [6, 61] and Statistics Canada JVWS [10, 62]. These are employer or establishment surveys, with administrative registers used as sampling frames or for calibration; they count posts that the employer is actively trying to fill from outside by any channel, advertised online or not, at a reference date or over a quarter or month, weighted to the economy they cover, published with a lag and later revised [5, 63, 54]. JOA is a daily-updated snapshot of advertised postings, so levels differ by design, as they do for other advert-based sources (one UK job board's stock of vacancies was around 40% of the ONS survey level [64]). What should be compared is structure, which countries, states and provinces are large, and, as the record lengthens, trends; sections 2 and 3 report that consistency check.

- **Level.** JOA's open postings are equivalent to 6.2% to 19.9% of the official counts across 14 countries. The ratio compares two different counts; it is not a coverage rate, and a value above or below another country's says nothing about data quality on its own.
- **Rank order.** The rank correlation of 0.93 across 14 countries mostly reflects country size, with 14 data points; the US state comparison (0.96 across 51 jurisdictions) uses JOLTS state values for December 2025, about nine months before the snapshot.
- **Country not resolved.** 26.5% of open postings (560,385) have no resolved country; they count in totals and are left out of country, state and province figures. The share ranges from under 1% for several recruiting systems to over 80% for four others, so each country's ratio is understated by a different amount; spread in proportion, the unknown-country postings would raise every ratio by about 36%.
- **Sweden.** 33% of Sweden's open postings (7,981) come from the Swedish public employment service's feed; without them Sweden's ratio is 13.2%.
- **Timing.** JOA is a one-day snapshot; most official figures are for the second quarter of 2026, some are not seasonally adjusted, and seasonal movements between quarters can reorder countries with similar ratios.

### How pay, work model and role family are measured

- **Pay** is read from structured fields of the source and, for Austria and Germany, also from German advert text, where JOA labels it 'inferred' in its API. The text reader takes the first euro amount next to a pay word and records a single (minimum) amount with an hourly, monthly or yearly period. The report uses whether a pay figure is present, not its amount; the sampled estimate below shows how far detection falls short of the pay that postings state.
- **Pay verification sample.** To estimate how often postings actually state pay, two AI agents with web access read random samples of Austrian and German postings, blind to JOA's flags, in three groups (JOA detected pay; no detection but a pay-like phrase in the text; neither). The samples were drawn on 3 October; 16 Austrian and 10 German sampled postings no longer in the 4 October population with the same country were dropped, leaving 84 and 100, from 68 and 91 employers. Each posting was read on its live page where still open (61 Austrian and 83 German) and otherwise from JOA's stored text (23 and 17), with a quoted passage that a script checked against the text. Readers confirmed pay in 28 of 28 Austrian and 24 of 29 German postings where JOA had detected it; they found pay in 35 of 36 and 24 of 39 postings with a pay-like phrase JOA had not read, and in 8 of 20 and 1 of 32 postings with neither, mostly figures shown on the live page outside the stored description. A second, rule-based detector run on the same texts agreed with the readers on 97.6% of Austrian and 91.0% of German postings (kappa 0.91 and 0.82). Weighting the groups by their size in the population gives 77.9% for Austria (95% interval 69.8% to 85.8%) and 17.7% for Germany (12.8% to 26.2%), against 33.7% and 4.9% detected; the ratio is 4.4 (interval 2.9 to 6.2) against 6.9 detected. The 'neither' group carries much weight: in Austria (8 of 20 stating pay) it contributes about 14 points of the 78%; in Germany (1 of 32) the upper end of its interval would add about 12 points. Intervals reflect sampling and clustering by employer, not reader error.
- **Pay amounts check.** A separate AI-assisted reading of 60 Austrian and 60 German postings with an amount read from text found strictly correct amounts in 57 of 60 and 43 of 60; about 38% of the Austrian amounts are collective-agreement minima. No pay level is published in this edition.
- **Work model** is counted only when the posting states it: 9% of North American and 15% of European postings. JOA's inferred work model is a separate API field, not used here.
- **Role family** is classified from the job title by JOA's model; it is a classification of the title, not a field the employer wrote.

### Location

Country, state and province come from the source's location data and the posting's own page. Blind samples of 150 postings each, read on the live page or stored text by an AI reader who did not see the coded country, put the share of postings located in another country than the one coded at 6.0% for Germany (9 of 150; 95% interval 3.2% to 11.0%; sample drawn after the location correction of 4 October 2026) and about 6% for Austria (3% to 11%; 13 of 150 before that correction, projected for the applied rule). Postings without a resolved location come mostly from a few recruiting systems (UKG 100.0%, Personio 94.9%, Greenhouse 88.0%, Pinpoint 80.1% unresolved), so those systems are under-represented in the country figures for pay, posting dates and system mix; Personio, for example, has the highest share of stated dates more than 90 days old (65.6%). 83.3% of Canadian postings and 205,370 US postings carry no province or state.

### Stock figures, posting dates and the weekly panel

JOA's record starts on 1 July 2026 (employers' own careers pages from 2 August). Closure dates record when JOA's checks confirmed a closure, so this edition reports stock figures at the snapshot date and no closure, duration or flow figures. Published evidence on how long postings stay online comes from single platforms: a median posting duration of seven days on one technical-jobs board [48], and theory and a model-based estimate that older listings are less likely to be live vacancies [65, 66]. A stated posting date more than 90 days old describes the posting, not whether the role is still being filled.

The weekly panel (section 6.4) counts postings by employer-stated posting date from employers that JOA had seen by 20 July 2026 on seven recruiting systems with consistent posting dates (SmartRecruiters, Greenhouse, Ashby, Lever, Workable, Teamtailor, Breezy), in complete Monday-to-Sunday weeks from 31 August 2026. It includes postings that have since closed and, being fixed, covers established employers only: an employer that changed recruiting system after 20 July counts as a decline. Adding the Workday panel, whose weekly volumes are more volatile, the series move from 10,067 to 11,201 in Europe (+11.3%) and from 53,400 to 45,699 in North America (-14.4%); North American Workday volume fell by 58% in the week of Labor Day. Posting dates are date-only; a posting dated at a weekend may have been entered earlier.

### Definitions and thresholds

**Definitions used in this report**

| Term | Definition |
| :--- | :--- |
| Open posting | A job posting (advert) that JOA's daily check found live at the employer's own recruiting system or careers page, not opted out by the employer, passing JOA's quality checks and served by JOA's API. |
| Employer-direct | Published by the employer on its own applicant-tracking system or careers page, or received through a public employment service feed (a small share, mainly in Sweden). |
| Recruiting system | Software employers use to publish and manage postings (applicant-tracking system, ATS), e.g. Workday or SmartRecruiters. |
| Structured pay field | Pay published in a dedicated data field of the source, rather than only in the posting text. |
| Employer | A distinct company record in JOA. A corporate group can appear as several records; one company can appear under several countries. |
| Country | The country JOA resolves for the posting from the source's location data and the posting's own page. |
| Own careers pages | Postings read from the HTML of employers' own careers pages rather than from an applicant-tracking system's interface; about half of open postings. |
| Pay figure | A positive amount of at most 1,000,000 euro a year after conversion, either (a) published in a structured field of the source (lower bound of the range) or (b) read by JOA from German advert text (Austria and Germany only). Detected by JOA's automated reading; see the sampled estimate in this section. |
| Work model | Remote, hybrid or on-site, counted only when the employer states it. |
| Role family | Classified by JOA from the job title; published at a confidence of 0.6 or higher, the rest counted as unclassified. |
| Posting date | The date stated by the employer or its system, to the day. Systems define it differently (first publication, last update, start date), so its age is most comparable within one system. |
| Region | Europe: EU27, Switzerland, Norway, Iceland, Liechtenstein, United Kingdom. North America: United States and Canada. JOA's website also counts Mexico in North America, so its regional totals on /coverage and /public/site-stats are larger. |
| Thresholds | A percentage is shown only if its base has at least 100 postings; a pay level needs at least 30 stated values (none is published here). Chart minimums: 150 Austrian postings per recruiting system (5.2), 2,500 own-careers-page postings (5.3), 4,000 own-careers-page postings per US state (5.4), 1,000 dated postings (6.1), 5,000 open postings (6.2), 1,000 postings stating a work model (6.3). Weekly-panel movements under 10% over four weeks are not read as trends; this is a judgement rule, not a statistical test. |

### Reproducibility

The report is generated by a script from one read-only snapshot: it computes every figure from the snapshot's aggregate tables, draws the charts and writes the CSV. Automated checks trace every number of 1,000 or more in the text to a data file and test each chart against its stated selection rule. The downloadable CSV contains every aggregate behind the charts and tables with its sample size, and the data dictionary defines each column. Counts by country can be compared with JOA's keyless endpoint api.jobopportunitiesapi.org/public/facets, which reflects the live population at the time of the request. The query logic is available on request (section 9).

## 8. Next edition

Future editions extend this snapshot into a running view of the market.

- **Flows and duration.** Monthly new postings and closures on a stable panel, net change in the stock, and how long roles stay open by country and role family, once a full month of daily-verified closure dates is available.
- **Change over time.** Month-on-month comparison of the measures in this edition, with intervals.
- **Pay.** Pay levels and ranges by country and role family, and pay detection in more languages, starting with English, French and Dutch postings.
- **Coverage.** More recruiting systems and public employment services that allow redistribution, with priority for the markets where JOA's postings are smallest relative to official counts (Austria, Canada, Poland and Greece), province detail for Canada, and the effect of each new source reported separately.
- **Benchmark.** An updated comparison with official vacancy statistics in every edition, so readers can follow how JOA's counts move against them.

## 9. About this report

> **Note: About this report**
>
> Job Opportunities API (JOA, jobopportunitiesapi.org) is a one-person business registered in Greece and run by Loukas (Luca) Tzekos, who is editorially responsible for this report. JOA collects job postings from employers' own recruiting systems and careers pages and from a few public employment service feeds, re-checks every posting daily, honours employer opt-outs, and sells access to the data through an API, so it has a commercial interest in the data being used. The report is free and uses only JOA's own data and the public sources cited. Nobody commissioned or paid for it: no employer, software vendor or public body funded or reviewed it. AI tools helped prepare the analysis and text and ran the checks; Loukas (Luca) Tzekos reviewed and approved them (see jobopportunitiesapi.org/ai). Contact: hello@jobopportunitiesapi.org.

**Cite as:** Job Opportunities API (2026). JOA Job Market Report, September 2026 edition (snapshot of 4 October 2026). https://jobopportunitiesapi.org/reports/job-market-september-2026. Licensed under CC BY 4.0.

**Licence.** This report, its charts and its aggregate data are licensed under the Creative Commons Attribution 4.0 International licence (CC BY 4.0): share and adapt them, including commercially, with credit and a link to the report. Every chart has an embed snippet that includes the credit link. Third-party figures quoted here (official statistics; Indeed Hiring Lab data, CC BY 4.0) remain under their publishers' terms; credit them as cited.

**Data downloads.** The CSV holds every aggregate behind the charts and tables with its sample size, and the data dictionary defines each column; both contain aggregates only, with no postings, employer names or personal data. Row-level, provenance-tagged postings are available through the JOA API (jobopportunitiesapi.org/pricing).

**Corrections.** To report an error, a figure that fails to reproduce or a comparison you think is misleading, write to hello@jobopportunitiesapi.org with the subject 'Market report correction'. Each correction is checked, fixed on this page and listed below with its date. Researchers who want the query logic or a different cut of the aggregates can ask the same way.

### Corrections log

No corrections since publication.

## Audio chapters

1. Summary (1 minute 53 seconds): https://jobopportunitiesapi.org/reports/job-market-september-2026/audio/s00.mp3
2. 1. Key findings (3 minutes 38 seconds): https://jobopportunitiesapi.org/reports/job-market-september-2026/audio/s01.mp3
3. 2. Europe: overview (3 minutes 40 seconds): https://jobopportunitiesapi.org/reports/job-market-september-2026/audio/s02.mp3
4. 3. North America: overview (3 minutes 4 seconds): https://jobopportunitiesapi.org/reports/job-market-september-2026/audio/s03.mp3
5. 4. Country and state deep dives (5 minutes 11 seconds): https://jobopportunitiesapi.org/reports/job-market-september-2026/audio/s04.mp3
6. 5. Pay transparency: the law and the postings (8 minutes 1 second): https://jobopportunitiesapi.org/reports/job-market-september-2026/audio/s05.mp3
7. 6. Hiring dynamics (4 minutes 15 seconds): https://jobopportunitiesapi.org/reports/job-market-september-2026/audio/s06.mp3
8. 7. Methodology and limitations (13 minutes 26 seconds): https://jobopportunitiesapi.org/reports/job-market-september-2026/audio/s07.mp3
9. 8. Next edition (58 seconds): https://jobopportunitiesapi.org/reports/job-market-september-2026/audio/s08.mp3
10. 9. About this report (1 minute 30 seconds): https://jobopportunitiesapi.org/reports/job-market-september-2026/audio/s09.mp3

Narrated with a synthetic voice (text-to-speech). Tables, charts and references are not read.

## References

1. Eurostat. Job vacancy statistics, quarterly (dataset jvs_q_r21, NACE Rev. 2.1, updated 15 September 2026; CH from jvs_q_nace2, Q4 2025). <https://ec.europa.eu/eurostat/api/dissemination/statistics/1.0/data/jvs_q_r21> Accessed 2026-10-03.
2. Institut fuer Arbeitsmarkt- und Berufsforschung (IAB). IAB-Stellenerhebung, second quarter 2026: about 1.02 million open positions. <https://iab.de/presseinfo/iab-stellenerhebung-rueckgang-der-offenen-stellen-im-zweiten-quartal-2026-auf-rund-1-million/> Accessed 2026-10-03.
3. UK Office for National Statistics. Vacancies and jobs in the UK: September 2026, vacancy survey, total vacancies (thousands, seasonally adjusted, 3-month average June-August 2026). <https://www.ons.gov.uk/employmentandlabourmarket/peopleinwork/employmentandemployeetypes/timeseries/ap2y/unem> Accessed 2026-10-03.
4. Office for National Statistics (2021, last revised 14 June 2021). Using Adzuna data to derive an indicator of weekly vacancies: Experimental Statistics. URL https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/conditionsanddiseases/methodologies/usingadzunadatatoderiveanindicatorofweeklyvacanciesexperimentalstatistics Accessed 2026-10-03.
5. Eurostat. Job vacancy statistics (jvs), reference metadata (ESMS). https://ec.europa.eu/eurostat/cache/metadata/en/jvs_esms.htm Accessed 2026-10-04.
6. US Bureau of Labor Statistics. Job Openings and Labor Turnover Survey (JOLTS): job openings, total nonfarm, seasonally adjusted, August 2026 (preliminary), and state estimates (latest returned: December 2025). Retrieved through the BLS public API. <https://www.bls.gov/jlt/> Accessed 2026-10-03.
7. Instituto Nacional de Estadistica (Spain). Encuesta Trimestral de Coste Laboral, job vacancies, table 6047, Q2 2026. <https://servicios.ine.es/wstempus/js/EN/DATOS_TABLA/6047> Accessed 2026-10-03.
8. DARES (France). Emplois vacants, Q2 2026 (Acemo quarterly survey; all-size estimate and firms with 10+ employees). data.dares.travail-emploi.gouv.fr. <https://data.dares.travail-emploi.gouv.fr> Accessed 2026-10-03.
9. Hellenic Statistical Authority (ELSTAT). Job Vacancies, Q2 2026 (press release of 15 September 2026). <https://www.statistics.gr/en/statistics/-/publication/SJO41/2026-Q2> Accessed 2026-10-03.
10. Statistics Canada. Job vacancies, payroll employees, and job vacancy rate, tables 14-10-0432-01 (seasonally adjusted) and 14-10-0371-01 (provinces, not seasonally adjusted), July 2026. <https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=1410043201> Accessed 2026-10-03.
11. Statistik Austria. Offene Stellen (job vacancies), Q2 2026. <https://www.statistik.at/statistiken/arbeitsmarkt/arbeitskraeftenachfrage/offene-stellen> Accessed 2026-10-03.
12. ISTAT. Il mercato del lavoro, II trimestre 2026 (job vacancy rate 1.5%; no vacancy count published). <https://www.istat.it/comunicato-stampa/il-mercato-del-lavoro-ii-trimestre-2026/> Accessed 2026-10-03.
13. European Parliament and Council. Directive (EU) 2023/970 of 10 May 2023 to strengthen the application of the principle of equal pay for equal work or work of equal value between men and women through pay transparency and enforcement mechanisms (Arts 5, 7, 9, 34). EUR-Lex, CELEX 32023L0970. <https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX:32023L0970> Accessed 2026-10-03.
14. EUR-Lex. National transposition measures communicated for Directive (EU) 2023/970 (CELEX 32023L0970); counts of communicated measures, not proof of full transposition. Snapshot of 3 Oct 2026. <https://eur-lex.europa.eu/legal-content/EN/NIM/?uri=CELEX:32023L0970> Accessed 2026-10-03.
15. Austria. Gleichbehandlungsgesetz (GlBG), sections 9 and 10, consolidated version. Rechtsinformationssystem des Bundes (RIS). <https://www.ris.bka.gv.at/GeltendeFassung.wxe?Abfrage=Bundesnormen&Gesetzesnummer=20003395> Accessed 2026-10-03.
16. Poland. Labour Code art. 18(3ca) as amended by the Act of 4 June 2025 (Dz.U. 2025 item 807). Sejm ELI. <https://api.sejm.gov.pl/eli/acts/DU/2025/807> Accessed 2026-10-03.
17. Germany. Entgelttransparenzgesetz (EntgTranspG), sections 5 and 10. Gesetze im Internet. <https://www.gesetze-im-internet.de/entgtranspg/index.html> Accessed 2026-10-03.
18. Spain. Real Decreto 902/2020 on equal pay for women and men. Boletin Oficial del Estado BOE-A-2020-12215. <https://www.boe.es/buscar/act.php?id=BOE-A-2020-12215> Accessed 2026-10-03.
19. United Kingdom. Equality Act 2010, section 77. legislation.gov.uk. <https://www.legislation.gov.uk/ukpga/2010/15/section/77> Accessed 2026-10-03.
20. Netherlands. Bill 36949, implementation of the pay transparency directive (pending). Tweede Kamer. <https://gegevensmagazijn.tweedekamer.nl/> Accessed 2026-10-03.
21. California Labor Code section 432.3 (SB 1162, effective 1 January 2023). Legislative Information. <https://leginfo.legislature.ca.gov/faces/codes_displaySection.xhtml?lawCode=LAB&sectionNum=432.3> Accessed 2026-10-03.
22. Illinois Equal Pay Act of 2003, 820 ILCS 112/10(b-25) (P.A. 103-539, effective 1 January 2025). Illinois General Assembly. <https://www.ilga.gov/legislation/ilcs/fulltext.asp?DocName=082001120K10> Accessed 2026-10-03.
23. Maryland Labor and Employment Article section 3-304.2 (HB 649 / SB 525 of 2024, effective 1 October 2024). Maryland General Assembly. <https://mgaleg.maryland.gov/mgawebsite/Laws/StatuteText?article=gle&section=3-304.2&enactments=false> Accessed 2026-10-03.
24. Massachusetts General Laws c. 149 section 105F (Acts 2024, c. 141). Massachusetts Legislature. <https://malegislature.gov/Laws/SessionLaws/Acts/2024/Chapter141> Accessed 2026-10-03.
25. Minnesota Statutes section 181.173 (2024 c. 110 art. 7, effective 1 January 2025). Office of the Revisor of Statutes. <https://www.revisor.mn.gov/statutes/cite/181.173> Accessed 2026-10-03.
26. New Jersey N.J.S.A. 34:6B-23 (P.L. 2024, c. 91), in effect 1 June 2025. NJ Department of Labor. <https://www.nj.gov/labor/myworkrights/wages/pay-transparency/> Accessed 2026-10-03.
27. New York Labor Law section 194-b (pay transparency), effective 17 September 2023. NYS Department of Labor. <https://dol.ny.gov/pay-transparency> Accessed 2026-10-03.
28. Vermont 21 V.S.A. section 495o (Act 155 of 2024, H.704), effective 1 July 2025. Vermont General Assembly. <https://legislature.vermont.gov/bill/status/2024/H.704> Accessed 2026-10-03.
29. Revised Code of Washington 49.58.110 (effective 1 January 2023). Washington State Legislature. <https://app.leg.wa.gov/RCW/default.aspx?cite=49.58.110> Accessed 2026-10-03.
30. District of Columbia Code section 32-1453.01 (D.C. Law 25-138, applicable from 30 June 2024). DC Council. <https://code.dccouncil.gov/us/dc/council/code/sections/32-1453.01> Accessed 2026-10-03.
31. British Columbia. Pay Transparency Act, SBC 2023, c. 18, section 2. BC Laws. <https://www.bclaws.gov.bc.ca/civix/document/id/complete/statreg/23018> Accessed 2026-10-03.
32. Ontario. Employment Standards Act, 2000, Part III.1 (pay transparency in job postings). e-Laws Ontario. <https://www.ontario.ca/laws/statute/00e41> Accessed 2026-10-03.
33. Delaware Code Title 19, section 709C (marked effective 26 September 2027). Delaware Code Online. <https://delcode.delaware.gov/title19/c007/sc01/index.html> Accessed 2026-10-03.
34. Connecticut General Statutes section 31-40z (wage range on request or at offer stage). Connecticut General Assembly. <https://www.cga.ct.gov/current/pub/chap_557.htm> Accessed 2026-10-03.
35. Colorado General Assembly. SB19-085 Equal Pay For Equal Work Act, bill summary page (statute section and effective date not verified by JOA). <https://leg.colorado.gov/bills/sb19-085> Accessed 2026-10-03.
36. Indeed Hiring Lab. Pay Transparency Tracker data (pay-transparency-country.csv). URL https://github.com/hiring-lab/pay-transparency (raw file https://raw.githubusercontent.com/hiring-lab/pay-transparency/master/pay-transparency-country.csv; repository last updated by commit 2026-09-16) Accessed 2026-10-03.
37. Batra, H., Michaud, A. and Mongey, S. (2023). Online Job Posts Contain Very Little Wage Information. NBER Working Paper 31984. DOI 10.3386/w31984. URL https://www.nber.org/papers/w31984 Accessed 2026-10-03.
38. Arnold, D., Quach, S. and Taska, B. (2025). The Impact of Pay Transparency in Job Postings on the Labor Market. NBER Working Paper 34480 (17 Nov 2025). DOI 10.3386/w34480. URL https://www.nber.org/papers/w34480 Accessed 2026-10-03.
39. Indeed Hiring Lab. Remote Tracker data (remote_postings.csv). URL https://github.com/hiring-lab/remote-tracker Accessed 2026-10-03.
40. Hansen, S., Lambert, P. J., Bloom, N., Davis, S. J., Sadun, R. and Taska, B. (2023, revised Jan 2026). Remote Work across Jobs, Companies, and Space. NBER Working Paper 31007. DOI 10.3386/w31007. URL https://www.nber.org/papers/w31007 Accessed 2026-10-03.
41. Adrjan, P., Ciminelli, G., Judes, A., Koelle, M., Schwellnus, C. and Sinclair, T. (2021). Will it stay or will it go? Analysing developments in telework during COVID-19 using online job postings data. OECD Productivity Working Papers No. 30. DOI 10.1787/aed3816e-en. Record: https://econpapers.repec.org/RePEc:oec:ecoaac:30-en Accessed 2026-10-03.
42. Cedefop (2019). Online job vacancies and skills analysis: a Cedefop pan-European approach. Luxembourg: Publications Office of the European Union. https://www.cedefop.europa.eu/files/4172_en.pdf (doi 10.2801/097022 as printed in the report) Accessed 2026-10-04.
43. Statistics Netherlands (CBS). Online vacancies as a source for statistics (project page, text in Dutch; title translated). https://www.cbs.nl/over-ons/onderzoek-en-innovatie/project/online-vacatures-als-bron-voor-statistiek Accessed 2026-10-04.
44. Eurostat. Online job advertisement rate - Experimental statistics (OJAR). URL https://ec.europa.eu/eurostat/web/experimental-statistics/online-job-advertisement-rate Accessed 2026-10-03.
45. Lightcast Knowledge Base (24 May 2023). How does Lightcast handle duplicate postings? https://kb.lightcast.io/en/articles/6957661-how-does-lightcast-handle-duplicate-postings Accessed 2026-10-04.
46. Lightcast Service Status (notice of 17-18 Nov 2023). UK Postings - Implementing deduplication improvements. https://status.lightcast.io/notices/q3e6wtxwotxrmat6-uk-postings-implementing-deduplication-improvements Accessed 2026-10-04.
47. Indeed Hiring Lab. Indeed Data Frequently Asked Questions. https://hiringlab.indeed.com/indeed-data-faq-2/ Accessed 2026-10-04.
48. Davis, S. J. and Samaniego de la Parra, B. (2024, revised Dec 2025). Application Flows. NBER Working Paper 32320. DOI 10.3386/w32320. URL https://www.nber.org/papers/w32320 Accessed 2026-10-03.
49. Beręsewicz, M. and Pater, R. (2021). Inferring job vacancies from online job advertisements, 2021 edition. Eurostat Statistical Working Papers, KS-TC-20-008. Publications Office of the European Union. https://ec.europa.eu/eurostat/documents/3888793/12287170/KS-TC-20-008-EN-N.pdf/6a86d53e-d0b8-d608-988d-d91f0cef6c21?t=1611673495829 Accessed 2026-10-04.
50. Carnevale, A. P., Jayasundera, T. and Repnikov, D. (2014). Understanding Online Job Ads Data: A Technical Report. Georgetown University Center on Education and the Workforce, April 2014. URL https://cew.georgetown.edu/wp-content/uploads/2014/11/OCLM.Tech_.Web_.pdf Accessed 2026-10-03.
51. Hershbein, B. and Kahn, L. B. (2018). Do Recessions Accelerate Routine-Biased Technological Change? Evidence from Vacancy Postings. American Economic Review 108(7): 1737-1772. DOI 10.1257/aer.20161570. URL https://www.aeaweb.org/articles?id=10.1257/aer.20161570 Accessed 2026-10-03.
52. Cammeraat, E. and Squicciarini, M. (2021). Burning Glass Technologies data use in policy-relevant analysis: an occupation-level assessment. OECD Science, Technology and Industry Working Papers 2021/05. DOI 10.1787/cd75c3e7-en. https://ideas.repec.org/p/oec/stiaaa/2021-05-en.html Accessed 2026-10-04.
53. Kureková, L. M., Beblavý, M. and Thum-Thysen, A. (2015). Using online vacancies and web surveys to analyse the labour market: a methodological inquiry. IZA Journal of Labor Economics 4, article 18. DOI 10.1186/s40172-015-0034-4. Record: https://ideas.repec.org/a/spr/izalbr/v4y2015i1p1-2010.1186-s40172-015-0034-4.html Accessed 2026-10-03.
54. Office for National Statistics. Vacancy Survey QMI (quality and methodology information). https://www.ons.gov.uk/employmentandlabourmarket/peopleinwork/employmentandemployeetypes/methodologies/vacancysurveyqmi Accessed 2026-10-04.
55. Institut fuer Arbeitsmarkt- und Berufsforschung (IAB). IAB job vacancy survey (page in German, title translated). https://iab.de/das-iab/befragungen/iab-stellenerhebung/ Accessed 2026-10-04.
56. IAB Research Data Centre (FDZ). Data report 09/2025 on the IAB job vacancy survey (report in German, title translated). https://doku.iab.de/fdz/reporte/2025/DR_09-25.pdf Accessed 2026-10-04.
57. INSEE. Surveys on activity and employment conditions of the workforce (Acemo), source description (page in French, title translated). https://www.insee.fr/fr/metadonnees/source/serie/s1189 Accessed 2026-10-04.
58. Instituto Nacional de Estadistica (Spain). Quarterly Labour Cost Survey (ETCL), Q2 2026 press release with methodological note (in Spanish, title translated). https://ine.es/dyngs/Prensa/ETCL2T26.htm Accessed 2026-10-04.
59. ISTAT. Quarterly survey on job vacancies and hours worked (VELA), information on the survey. https://www.istat.it/informazioni-sulla-rilevazione/vela/ Accessed 2026-10-04.
60. Statistik Austria. Job vacancy survey (page in German, title translated). https://www.statistik.at/ueber-uns/erhebungen/unternehmen/offene-stellen-erhebung Accessed 2026-10-04.
61. US Bureau of Labor Statistics. Handbook of Methods, Job Openings and Labor Turnover Survey: design and data sources (Internet Archive copy of 2025). https://web.archive.org/web/2025/https://www.bls.gov/opub/hom/jlt/design.htm Accessed 2026-10-04.
62. Statistics Canada. Guide to the Job Vacancy and Wage Survey, 2024 (catalogue 75-514-G; marked archived, a more recent version exists). https://www150.statcan.gc.ca/n1/pub/75-514-g/75-514-g2024001-eng.htm Accessed 2026-10-04.
63. US Bureau of Labor Statistics. JOLTS data definitions: how does JOLTS define job openings (Internet Archive copy of 2025). https://web.archive.org/web/2025/https://www.bls.gov/jlt/jltdef.htm Accessed 2026-10-04.
64. Turrell, A., Speigner, B. J., Djumalieva, J., Copple, D. and Thurgood, J. (2019). Transforming Naturally Occurring Text Data Into Economic Statistics: The Case of Online Job Vacancy Postings. NBER Working Paper 25837. DOI 10.3386/w25837. URL https://www.nber.org/papers/w25837 Accessed 2026-10-03.
65. Albrecht, J., Decreuse, B. and Vroman, S. (2020). Directed Search with Phantom Vacancies. IZA Discussion Paper No. 13704. URL https://hdl.handle.net/10419/227231 Accessed 2026-10-03.
66. Ng, H. (2024). Why is it so hard to find a job now? Enter Ghost Jobs. arXiv:2410.21771 \[econ.GN\], submitted 29 Oct 2024. DOI 10.48550/arXiv.2410.21771. URL https://arxiv.org/abs/2410.21771 Accessed 2026-10-03.