TL;DR: The World Values Survey, Akaliyski et al.’s 2025 individualism-collectivism index, and the Minkov-Kaasa model provide more accurate, more current, and more granular cultural data than Hofstede’s 1960s IBM dataset. CultureFlow is built on all three.
If you have read why Hofstede’s cultural dimensions are outdated, the next question is practical: what should you use instead?
This is not a theoretical question. When a cross-cultural training programme, a hiring decision, or a pre-meeting briefing is built on data that overestimates American individualism by 27 points and underestimates Japanese individualism by 22 points, real decisions go wrong. The stakes of using better data are concrete.
Here are the three frameworks that have emerged as the credible alternatives – and what each one contributes that Hofstede cannot.
The World Values Survey: The Data Foundation Everything Else Builds On
The World Values Survey (WVS) is the most comprehensive cross-national research project into human values ever conducted. Since 1981, it has run seven waves of surveys across more than 100 countries, using nationally representative samples – not employees of a single company, not convenience samples, but actual cross-sections of national populations.
The WVS measures hundreds of values, beliefs, and attitudes: trust in institutions, attitudes to authority, gender equality norms, religiosity, work ethic, tolerance of outgroups, and much more. It is updated roughly every five years, meaning its data reflects cultural change over time rather than freezing a culture at a single historical moment.
Three things make WVS fundamentally more reliable than Hofstede as a data source:
Representativeness. Hofstede’s original data came from IBM employees in the late 1960s and early 1970s – a narrow, self-selecting sample of people who had all chosen to work for the same American corporation. WVS surveys are designed to be nationally representative across age, gender, education, urban/rural, and income.
Recency. The most recent WVS wave (Wave 7) covers 2017–2022. Hofstede’s foundational scores derive from data collected more than fifty years ago.
Breadth. WVS captures cultural complexity across hundreds of variables. Hofstede compressed culture into six dimensions. More dimensions means more signal and less distortion.
CultureFlow’s Cultural Intelligence Engine draws on 12,735 WVS graph edges across 100+ countries. When you receive a briefing, the cultural data underneath it comes from WVS Wave 7, not from IBM employee surveys.
Akaliyski et al. 2025: The Most Precise Individualism-Collectivism Index Available
For decades, the individualism-collectivism dimension has been the most widely applied in cross-cultural practice. It underpins assumptions about how people make decisions, respond to group pressure, assert their interests, and relate to authority. It is also the dimension where Hofstede’s biases are most severe.
A 2025 paper by Petko Akaliyski, Vivian Vignoles, Christian Welzel, and Michael Minkov, published in the Journal of Personality and Social Psychology, constructed a new individualism-collectivism index for 102 countries using WVS and European Values Study data. The methodology is directly comparable to Hofstede’s own, allowing a precise bias measurement.
The findings are stark.
Hofstede’s scores overestimate individualism in English-speaking countries by approximately 27 points on a 100-point scale. The United States, United Kingdom, Australia, and Canada all appear significantly more individualistic in Hofstede’s framework than values-based evidence supports.
Conversely, Hofstede underestimates individualism in East Asian countries by approximately 22 points. Japan, South Korea, and China all appear significantly more collectivist in Hofstede’s data than the WVS evidence supports.
These are not methodological quibbles. A 27-point bias changes how a culture is characterised. When a cross-cultural training programme tells participants that “Japanese culture is highly collectivist,” it is perpetuating a characterisation that the 2025 research shows to be measurably inaccurate.
The Akaliyski index is now incorporated into CultureFlow’s Cultural Intelligence Engine as the primary measure for individualism-collectivism, replacing Hofstede’s scores entirely for this dimension. All 101 country scores from the 2025 paper are ingested and active.
Minkov-Kaasa: Recalibrating Hofstede’s Own Framework
Michael Minkov’s contribution to cultural research is distinctive. Unlike researchers who abandoned Hofstede entirely, Minkov’s approach has been to test Hofstede’s framework using Hofstede’s own methodology, then compare the results to what the original data showed.
The Minkov-Kaasa model – developed with Anneli Kaasa and drawing on WVS data – provides updated country-level scores derived through Hofstede’s dimensional approach but grounded in nationally representative survey data rather than the IBM dataset. Where the two sets of scores diverge, the WVS-derived scores are consistently more aligned with observed cultural behaviour in modern research.
What makes Minkov’s work particularly valuable for applied cultural intelligence is its rigour about what culture is and is not. Minkov has been explicit that culture changes over time, varies within countries, and cannot be adequately captured by a static dimensional score. His research tracks cultural change across WVS waves, showing which dimensions are stable and which are shifting rapidly. This matters for practitioners: a country’s score from 1981 and its score from 2022 are not the same number, and acting as if they were produces errors.
CultureFlow uses the Minkov-Kaasa model as a calibration layer for country-level scores, validating cultural dimension outputs against the WVS-derived baseline and flagging where generational or urban-rural variation is particularly significant.
GLOBE: A Different Methodology, A Different Signal
The Global Leadership and Organizational Behavior Effectiveness (GLOBE) study, involving 17,000 managers across 62 societies, provides a third data source that complements rather than replaces WVS-based research.
GLOBE makes a distinction that Hofstede does not: the difference between a culture’s values (what people say they believe should be the case) and a culture’s practices (what people actually do). These are not always the same. A society can value egalitarianism while practising hierarchy. GLOBE measures both, and the gap between them is often where cultural friction lives.
For cross-cultural meeting preparation, GLOBE data adds signal on how cultures behave in organisational contexts specifically – which is more directly applicable than national-level values data for professional interactions.
What Changes When You Use Better Data
The practical differences are significant. Consider three examples.
Japan. Hofstede’s collectivism score for Japan is among the highest in the world. The Akaliyski 2025 index shows Japan as moderately collectivist, significantly less so than Hofstede suggests. In practice: Japanese professionals – particularly younger, internationally-facing ones – exercise considerably more individual agency than Hofstede-based frameworks predict. Training built on Hofstede’s characterisation may produce unnecessary deference and misread individual initiative as cultural deviation.
United States. Hofstede’s extreme individualism score for the US has shaped decades of cross-cultural theory. The Akaliyski data shows US individualism is overstated by 27 points. In practice: American professionals are more relationship-oriented and group-conscious than Hofstede suggests, particularly in organisational contexts. The “hyper-individual American” is partly a methodological artefact.
China. Hofstede shows China as highly collectivist. The WVS data shows this varies enormously by generation, urban-rural split, and industry. A 2022 graduate from Shanghai working in tech behaves very differently from what Hofstede’s single national score predicts. Using that score as a basis for cultural preparation is increasingly unreliable.
The Practical Upshot
Better data produces better preparation. When CultureFlow COMPASS generates a briefing for a meeting between a Norwegian seller and a Japanese buyer, it does not pull Hofstede’s 1960s scores. It draws on the Akaliyski 2025 index, the Minkov-Kaasa calibration model, the GLOBE organisational data, and 12,735 WVS edges to produce a briefing that reflects how cultures actually operate today.
This is not an academic distinction. The difference between Hofstede’s Japan score and the 2025 Akaliyski score represents a 22-point gap in how collectivist Japanese culture is characterised. That gap translates directly into how you interpret silence in a meeting, how you read consensus signals, and how much individual agency you expect from the people across the table.
The tools have caught up with the research. The question is whether the practitioners using them will too.