Between January and July 2026, an international schools group closed 749 background screening cases with Avvanz. Thirty-two cases came back with a confirmed discrepancy. Four percent reads as unremarkable until you convert it: that is one candidate in every twenty-three. And this for a top international schools group working with the best recruiters.
So a school screening sixty candidates across a recruitment season should expect two or three of them to present information about themselves that does not survive contact with the source. What makes the dataset useful is not the headline rate. It is the underlying pattern - which claims failed, and which checks caught them.
Discrepancy analysis: January to July 2026
| Check type | Discrepancy cases | Percentage | What surfaced |
|---|---|---|---|
| Criminal / Civil Litigation Checks | 7 | 22% | Civil matters, judgments and proceedings outside the criminal system. |
| Employment Checks | 15 | 47% | Dates, salary, title, rehire status and conduct or attendance concerns. |
| Education Checks | 7 | 22% | Attendance, graduation and qualification status. |
| Credit Checks | 2 | 6% | Discrepancies identified through credit screening. |
| Social Media Check | 1 | 3% | A discrepancy identified through public social media information. |
Cases closed: 749 Red cases: 32 Red percentage: 4%
Nearly half of all discrepancies found were about previous jobs
Fifteen of the thirty-two discrepancies, 47 percent, related to employment history. Five candidates had shifted or overstated their dates of employment. Four had inflated a salary against the payroll record. Three had given themselves a title more senior than the one they held.
Those three categories are accuracy failures, and a school can reasonably decide how much weight to give them. The remaining three cases are a different matter. Two candidates were marked not eligible for rehire by a former employer. One had conduct and attendance concerns raised by a referee during verification.
No CV will ever disclose either of those things. Neither will a reference letter the candidate supplied. They surfaced because an Avvanz team member went back to the previous employer independently and asked a direct question.
Education discrepancies were rarely fake degrees
Education claims accounted for seven findings, or 22 percent of the total. Almost none involved invented qualifications from institutions that do not exist - the scenario most screening conversations describe. What actually appeared was subtler.
Four candidates gave attendance or graduation dates that did not match the registrar's record, often closing an awkward gap in the timeline. Two presented a qualification as held when they were in fact only still enrolled. One undergraduate award could not be confirmed as described.
In a teaching post tied to a specific qualification, or in a jurisdiction where the work permit depends on the degree, "still studying" is not a paperwork issue. It is compliance exposure that lands on the school, not the candidate. The lesson is narrow and practical: a certificate is a picture of a claim. Verification means asking the awarding institution.
A significant source of red flags might surprise you
Seven findings - 22 percent - came from Criminal / Civil Litigation checks.
Civil matters sit outside the criminal system, so a clean police certificate says nothing about them. They surface unresolved disputes with former employers, judgments, and proceedings a candidate chose not to mention. Most school screening packages stop at identity, education and a criminal record check, which means this category is invisible by design in the majority of hiring processes.
Availability varies considerably across jurisdictions, and civil checks are not proportionate for every role. That is a legitimate constraint. But there is a real difference between a school that has decided, role by role, where civil litigation checks apply, and a school that has never considered the question.
Why international schools carry more of this risk
Three conditions compound in our sector. Careers cross borders, so a candidate arrives with employers and awarding bodies in three or four countries and a single-country check verifies only a fraction of the CV. Records sit in different systems, held to different rules, with wildly different turnaround. And hiring runs on a compressed calendar - recruitment fairs, resignation deadlines, competing offers - which makes verification the step most often shortened to hold a candidate.
None of this creates dishonesty. It reduces the chance that dishonesty is noticed.
Six things this data argues for
- Verify every employment episode across the last ten years, not only the most recent post.
- Ask the rehire question directly of each former employer, and record the answer.
- Confirm qualifications with the awarding institution rather than from a certificate.
- Set civil litigation scope by role and jurisdiction - and document the reasoning where it is excluded.
- Re-screen on promotion into safeguarding-critical and finance-facing roles.
- Keep one auditable evidence trail per hire, retained in line with your data protection policy.
Start with the second. It costs nothing, requires no service provider like Avvanz, and two of the thirty-two findings in this dataset came from it.
The unglamorous conclusion
In a school, a hiring decision and a safeguarding decision are the same decision. A teacher who moved a graduation date by a year is a records problem. A teacher who left a post under investigation and reappears three countries later is a child protection problem. Both arrive in the same envelope, on the same CV, and only verification distinguishes them.
Thirty-two cases raised findings that warranted closer review. Following up with employers, institutions and other sources takes time, but it gives schools a stronger basis for hiring decisions - helping them resolve discrepancies, investigate concerns and assess candidates beyond the information in their CV.