Opinions and values are not the same thing
Survey questions cover a wide range. At one end are views about a particular policy, politician or news event. These can shift quickly in response to events, campaigns and new information. At the other end are broad orientations: attitudes about family, religion, authority, fairness, community and personal freedom. These tend to be learned early, tied to identity and reinforced by the people around us, and they typically move far more slowly.
Mixing the two up is a common source of confusion in trend coverage. A sharp swing in views on a specific proposal does not by itself show that people’s underlying values have changed, and a slow drift in values can look like nothing at all in any single year.
Why deep attitudes are sticky
Several features of how people form views help explain the inertia.
- Early formation. Many researchers argue that core outlooks are shaped substantially in adolescence and early adulthood by family, schooling, community and the conditions of the time. After that, they often become relatively settled, though the extent varies between people and between topics.
- Identity and belonging. Values are often bound up with who a person is and which groups they belong to. Changing a view can carry a social cost, so people may resist evidence or arguments that challenge it.
- Selective exposure. People tend to spend time with others who share their outlook and to notice information that fits it, which reinforces what they already believe. How strongly this operates online is debated, and the evidence on filter bubbles is more mixed than popular accounts suggest.
- Consistency. Once a person has a coherent set of views, changing one can create tension with others, so change at the edges is easier than change at the core.
None of this means that individuals never change their minds. People do revise views after major life experiences, new circumstances or sustained exposure to different perspectives. But large, rapid shifts in individuals’ deep values are usually uncommon.
How whole societies change anyway
If individuals change slowly, how do societies end up with noticeably different attitudes after a few decades? A major part of the answer is generational replacement, sometimes called cohort replacement. Each generation grows up in somewhat different conditions, and its members carry their formative outlook with them as they age. Over time, older cohorts leave the population and younger ones enter. Even if nobody changed their mind, the overall average would move because the mix of people has changed.
Imagine, as a thought experiment, a society where each cohort stays fixed in its views for life, but successive cohorts are slightly more open to a certain idea than their predecessors. The national picture would shift steadily, with no one being converted. Real life is a blend: some change comes from replacement, and some from people updating their views, often in the same direction as the wider culture around them. Researchers try to estimate the share of each, and the answer depends on the topic and the period.
Age, period and cohort
When survey data show that younger people hold different views from older people, there are at least three possible explanations, and they can overlap.
- Age effects. People may hold different views at different stages of life, for example becoming more cautious with age, so a gap would reflect where people are in life rather than a lasting difference.
- Period effects. Events or shifts in the climate of opinion can move everyone at once, across all age groups.
- Cohort effects. Groups born at different times grow up under different circumstances and keep distinct outlooks.
These are hard to separate with a single survey, since age, period and cohort are mathematically linked. Repeated questions over many years help, which is one reason tracking surveys are valued. Even so, disentangling them often relies on assumptions, and researchers can disagree.
A related caution is that talk of generational labels can exaggerate differences. Sorting people into named generations is convenient, but differences within any generation are often larger than the differences between them.
Why surveys can mislead about speed
Two features of measurement can make change look faster or slower than it is.
First, a shift in what people are willing to say can mimic a shift in what they believe. As some views become more or less socially acceptable, answers may move ahead of private attitudes, an effect connected to social desirability. This is relevant to studies of changing social norms, where reported attitudes and actual attitudes may not move in step.
Second, a change in question wording or survey method can produce an apparent jump. A genuine change in deep values is usually gradual, so an abrupt break in a long series is a prompt to check the measurement before looking for a social explanation.
Fast and slow together
It helps to hold two ideas at once. Specific opinions can swing in weeks, especially on issues that are in the news. Underlying values typically evolve over decades, mostly through the arrival of new cohorts and, to a lesser degree, through gradual updating by existing ones. Stories that frame a short-term movement as a transformation of values often confuse the two layers.
Reading claims about changing values
A few questions are useful when a headline announces that attitudes have shifted:
- Is the question about a specific issue or a deep orientation?
- Is the comparison between the same people over time, or between different age groups at one time?
- Does the series use consistent wording across its whole length?
- Could willingness to express a view, rather than the view itself, have changed?
- Is the claim a description of change, or an explanation of why it happened?
Attitude change on values is real, but it is typically a slow, layered process. Approaching claims of sudden transformation with measured scepticism is usually the safer reading, while remaining open to the possibility that some changes, especially after major events, do happen faster.