SES (Socioeconomic Status)

You are running a study of consumer behavior. Knowing a respondent's income level is critical for your analysis, but the direct question "how much do you earn?" produces a low response rate and a lot of refusals.

Socioeconomic status (SES) is a multidimensional characteristic that includes income, education, occupation and several other indicators. Measuring through these turns out to be both more accurate and more acceptable to the respondent than a single question about money.

Definition

SES (Socioeconomic Status) is a composite indicator of the social position of a person or household, based on a combination of economic (income, assets), educational (level of education) and occupational (type of employment, job position) characteristics. It is used in sociological, marketing and epidemiological research for audience segmentation, analysis of social differences, targeting and understanding behavior. Individual SES indicators in surveys usually have a higher response rate than a direct question about income.

Components of SES

The classic SES model (after Duncan, Hollingshead and later adaptations) includes three core components:

Income. The level of personal or household income. It can be measured in absolute amounts (broken down into ranges: up to 50K, 50–100K, etc.) or relative to average values for the country/region. The most "sensitive" component — many people refuse to answer.

Education. The highest level of education attained: secondary, vocational, higher, postgraduate. A less sensitive category; most respondents answer without any problem.

Occupation. The type of employment and job position. Classifications are used (the international ISCO-08 and national equivalents). It allows status and income to be assessed indirectly.

Additional indicators that are often included in extended scales:

• Assets — owning a home, a car, a second/country home
• Area of residence — as a proxy for the socioeconomic environment
• Subjective assessment — how a person rates their own position relative to others (the MacArthur Subjective Social Status Ladder)

SES indices

Different research traditions use different ways of aggregating the components into a single indicator:

Hollingshead Index. A classic scale that combines education and occupation into a single score. Education is rated across 7 levels, occupation across 9 prestige categories. The final score is a weighted sum.

Duncan Socioeconomic Index (SEI). Based on occupation, but factoring in the income and education associated with it. Each occupational category has a fixed score that reflects the average "social position".

Index of Relative Socioeconomic Disadvantage. Used for geographic segments (neighborhood, postal code). It aggregates several indicators at the area level: income, employment, education, housing conditions.

Simple composites. In applied surveys a simple combination of 3–5 questions (income, education, employment, asset ownership) is often used, followed by categorization into 3–5 groups (low, lower-middle, middle, upper-middle, high SES).

How to include SES in surveys

A block of several indicators instead of just one. Instead of "how much do you earn", include 3–4 questions: level of education, field of employment, job position, home ownership. Each one on its own is less sensitive, and together they give a more reliable SES estimate.

An income question with ranges, not an exact amount. "Up to 30K / 30–60 / 60–100 / 100–200 / over 200" is easier to answer than typing in an exact figure. The ranges can be adapted to regional specifics.

Make the questions optional. Make SES questions optional for completing the survey — or offer a "prefer not to answer" option. This reduces the share of people who abandon the survey at this stage.

Position in the survey. Demographic and SES questions are usually placed at the end of the questionnaire. By that point the respondent has already invested time and answers more willingly than on the first screens.

Explaining the purpose. A short sentence — "These questions help us better understand the differences between groups of customers" — reduces suspicion and increases willingness to answer.

Example: SES segmentation in a consumption study

A study of online-service consumption among city residents. Four SES indicators were included in the survey:

• Level of education (5 options)
• Field of employment (10 options)
• Per-person household income (7 ranges)
• Home ownership

A simple composite index was built: each component was ranked from 1 to 5, for a total score from 4 to 20. Respondents were split into 3 groups:

• Low SES (4–9 points): 22% of the sample
• Middle SES (10–15 points): 58%
• High SES (16–20 points): 20%

Analysis of consumer preferences by group showed: the high-SES group is willing to pay for premium plans and values convenience, the middle group is sensitive to the price/quality ratio, and the low group is focused primarily on price. Without SES segmentation these differences would have blurred into the average figures across the whole sample.

Common mistakes when working with SES

Relying on income alone. Income is only one of the components. A retiree with a high level of education and an owned home in the city center may have a low current income but a high SES. A student from a well-off family has a low personal income but a high family SES. A single indicator oversimplifies the picture.

Using outdated categories. Classic occupational classifications may not account for new realities: the gig economy, remote work, the self-employed, the creative economy. Modern research needs adapted categories.

Ignoring the regional context. An income of a given amount in a capital city and in a small town represents different SES. In international or cross-regional research, income needs to be normalized to regional benchmarks or to use relative characteristics.

Social desirability in answers. Respondents may inflate their education or job position to look more prestigious. Social desirability bias works especially strongly in demographic questions. Ensuring anonymity reduces this distortion.

SES and ethical questions

Working with SES requires ethical care:

Respectful wording. Questions should not sound like "how poor are you". "Which range does your monthly income fall into" is neutral. "What is your highest level of education" is standard.

Not using it for discrimination. If SES data is going to be used for decisions about specific individuals (terms of service, prices, access to services), this can cross over into discrimination. For segmentation in research it is fine; for targeting at the individual level it requires care.

Confidentiality. SES data is part of personal data. It requires the same protection as other categories: informed consent, secure storage, the right to erasure.

SES in SurveyNinja

In SurveyNinja an SES block is created using standard single-choice questions. For composite indices it is convenient to use scoring: each answer option gets a score, and the total automatically determines the SES category. This lets you segment respondents right on the final screen or during analysis.

For correct SES analysis it is important to ensure the sample is representative on this parameter — otherwise the conclusions will be biased in favor of the dominant segment. For understanding the profile of the target audience, SES is one of the key parameters, alongside age, geography and gender.

SES is not a single question but a composite indicator made up of several indicators: income, education, occupation, assets. Measuring it through a block of questions gives a fuller and more accurate picture than a direct question about salary, and raises the response rate. For research tasks SES is a key segmentation parameter; for targeting at the individual level it requires ethical care.

Frequently asked questions

Which SES indicator is the most reliable?

Education is the most stable and reliable indicator: it rarely changes after it is attained, is not subject to short-term economic fluctuations, and respondents answer it willingly. Income is more sensitive to the current situation and more often triggers refusals. A combination of several indicators gives a more reliable estimate than any single one.

Do you need to include a direct income question?

It is desirable but not mandatory. If the task is rough SES segmentation, education and occupation are enough. If you need an accurate assessment of economic position, an income question is necessary. Make it optional and use ranges instead of exact amounts — this raises the response rate.

How do you adapt the SES scale for different countries?

Each country requires localization: regional specifics of income, classifications of occupations, education levels. Copying an American or European scale without adaptation produces incorrect results. International research uses cross-cultural scales (the ESOMAR Social Grade) or relative indicators.

Can SES be used for ad targeting?

In theory yes, but with ethical and legal limitations. In a number of jurisdictions (GDPR and others) targeting based on sensitive demographic parameters requires explicit consent. In sensitive areas (finance, health, housing) targeting by SES may be deemed discriminatory. For mass-market products it is usually acceptable; for critical ones it requires legal review.

What should you do if most respondents refuse to answer SES questions?

Analyze the reasons: the block is too long, the wording too direct, no explanation of the purpose, no anonymity, a poor placement in the questionnaire. What often helps: shortening the block, making the questions optional, explaining the purpose, moving them to the end of the questionnaire. If the refusal rate stays high, SES may not be critical for your task and you can do without it.