Table 2 Overview of variables used in the analysis, sorted alphabetically

age

Age of the respondent in years

1 (18–29), 2 (30–49) or 3 (50 or more)

colgknow

In respondent’s opinion, how many work colleagues or schoolmates know that respondent is gay

1 (none), 2 (a few), 3 (most) or 4 (all)

colgopen

To how many work colleagues or schoolmates is respondent open about being gay

1 (none), 2 (a few), 3 (most) or 4 (all)

concealment (only LRM)

Index of concealment of sexual orientation at work

Scale from 0 to 1, higher value stands for more concealment

discrexp

During 12 months preceding the survey, the respondent has personally felt discriminated against because of being gay at work

0 (no) and 1 (yes)

discrprev

Country-level index capturing the arithmetic average of answers of lesbians (gays) on how prevalent is discrimination against lesbians (gays) in their country. The respondents are assigned the value corresponding to their sex

Scale from 0 to 1, higher value stands for more better legal situation for gays and lesbians

education

The highest level of education that the respondent achieved

1 (primary or lower), 2 (secondary), 3 (post-secondary other than college/university), 4 (college/university/or higher)

expnegatt

During employment in the 5 years preceding the survey, the respondent experienced a general negative attitude at work against people because they are LGBT

1 (never), 2 (rarely), 3 (often) or 4 (always)

homonegativity (only LRM)

Index of homonegativity at work

Scale from 0 to 1, higher value stands for more homonegativity

legislation

Index on legal situation regarding equality and non-discrimination based on sexual orientation. It is calculated from ILGA Rainbow Index 2012 (ILGA Europe 2012) and it captures protection from discrimination by constitution, in employment, in goods and services, in other spheres of life, by equality body mandate or by equality action plan. The index is obtained by dividing the actual country score by the maximum score the country can achieve

Scale from 0 to 1, higher value stands for more better legal situation for gays and lesbians

negcondct

During employment in the 5 years preceding the survey, the respondent experienced negative comments or conduct at work because of you being gay

1 (never), 2 (rarely), 3 (often) or 4 (always)

openfear

The respondent avoids being open about being gay in the workplace for fear of being assaulted, threatened or harassed by others

0 (no) and 1 (yes)

percdiscr (only LRM)

Index of perceived discrimination at work

Scale from 0 to 1, higher value stands for more perceived discrimination

reporting

The most recent discrimination incident was reported by the respondent or anyone else

0 (no) and 1 (yes)

unemployed

Variable capturing whether a respondent’s current employment status is unemployed

0 (no) and 1 (yes)

unemployment rate

The annual average unemployment rate per country in 2012, based on variable une_rt_a from Labour Force Survey (Eurostat 2017). The separate unemployment rates per sex are not used because the labour market attachment of lesbians (gays) partly resembles the attachment of heterosexual men (women) (Fric 2017)

% of unemployed people on active population

witcondct

During employment in the 5 years preceding the survey, the respondent heard or seen negative comments or conduct because a colleague is perceived to be LGBT

1 (never), 2 (rarely), 3 (often) or 4 (always)

workhide

During employment in the 5 years preceding the survey, the respondent has hidden or disguised being gay at work

1 (never), 2 (rarely), 3 (often) or 4 (always)

workopen

During employment in the 5 years preceding the survey, the respondent has been open about you being gay at work

1 (never), 2 (rarely), 3 (often) or 4 (always)