EDI Reality Check

Start with what the law already requires, then compare outcomes you lawfully hold and how people say they experience the workplace, so you judge progress by effect instead of by the number of initiatives.

Inclusion theory is well developed (Shore et al., 2011); the effects of diversity training and specific programs vary, and accountability has done more than training in the best known study (Kalev, Dobbin & Kelly, 2006; meta-analysis: Bezrukova et al., 2016). Psychological safety is strong research (Edmondson, 1999). Culture fit can slide into cultural matching (Rivera, 2012). Legal references are to the Swedish Discrimination Act, the EU pay transparency directive, GDPR and the German Works Constitution Act. The stage scale in the optional part is a self-assessment made for this page, not a research model.

Use it when:

• You lead a team or organization and want an honest picture before you set equality goals or plans.
• Your active measures or pay survey are due, and you want them to lead to change rather than paperwork.
• There are many diversity activities but no one can say what they changed.

Skip it when:

• You want practical steps to make daily work more inclusive in one team: use 4.18 Inclusive Workplace Checklist.
• The question is one hiring process: use 4.15 Recruitment Funnel Map and 4.16 Selection Quality Scorecard.

The evidence on equality, diversity and inclusion is mixed, and this check is built for that. The concept is well developed: Shore et al. (2011) describe inclusion as people both belonging and being valued for what makes them distinct, which is what the climate statements here ask about. What varies is the effect of specific programs. A meta-analysis of diversity training found effects that differed by outcome and faded for some (Bezrukova et al., 2016), and in a long study of workplaces, making someone accountable did more for diversity in management than training did (Kalev, Dobbin & Kelly, 2006). So counting initiatives proves little; the check keeps asking for evidence of effect. Two further findings shape it. Rivera (2012) showed how hiring for fit turns into hiring people like yourself, which is why fit appears as a barrier. And where speaking up feels unsafe, problems stay hidden and every other score looks better than it is (Edmondson, 1999).

The legal duties come first because they are the floor, not the ambition. The limit: in a small team the numbers are too small to show patterns, and identity data you may not lawfully hold cannot be filled in.

In Sweden, all employers must work continuously on active measures, together with employees, in four steps: investigate, analyze, act and follow up. The work covers all seven grounds of discrimination (sex, gender identity or expression, ethnicity, religion or other belief, disability, sexual orientation and age), by looking at policies, processes and anonymized survey results, not by counting who belongs to which group. Outside Sweden, use the equivalent national rules.

Strong answer: "Recruitment: investigated in March, risk found that all interviews are run by one manager; action: two interviewers and a scoring guide from the autumn, owner HR partner; follow-up due in December."

Biggest risk found, the action, its owner and date:

Pay survey this year (date, unjustified differences found and corrected; written documentation is required from 10 employees, and for all active-measures work from 25):

Ready to give applicants and employees pay information under the pay transparency directive? (yes, partly, no; see the box below)

Sources: DO: active measures for employers, DO e-guide: pay survey.

Illustrative example: the roles, the organization and all numbers are made up.

The HR manager at a municipal housing company in central Sweden, 80 employees, runs the check before the annual active measures review.

Section 1: recruitment and pay investigated and acted on; working conditions and parenthood only investigated; skills development not started. Pay survey done in February, two unjustified differences corrected.

Section 2: women are about half of staff but one in four managers; promotion rates differ more than pay. The gap opens at the step into first-line management.

Section 3: 58 asked, 41 answered. Lowest statement: confidence in reporting, clearly below the rest.

Section 4: two rounds of bias training with good feedback and no change in the numbers. Biggest barrier: acting manager roles are handed out informally.

Section 5: "We think promotion is open, and it is not." Actions: acting and stretch roles advertised internally with criteria, owner HR manager, from next quarter; a clear reporting route with a named receiver and a promised response time, owner CEO with the union, by June. Follow-up planned in the next active measures cycle.