Hofstede and GLOBE scores for Vietnam: hypotheses, not answers

GLOBE's list of surveyed countries runs from Albania to Zimbabwe, and Vietnam is not on it. Hofstede's scores do cover Vietnam, but the page of The Culture Factor Group (formerly Hofstede Insights) that I opened on 3 November 2026 gives individualism as 30 and long-term orientation as 47, where an archived copy of 1 July 2022 gave 20 and 57. The page gives only its latest update, 16 October 2023, for those two dimensions, and does not say why. If you are about to change how 45 people work and want national culture to shape the design, how much weight can these numbers carry? This post looks at how to use Hofstede's six cultural dimensions and GLOBE's nine cultural dimensions to form hypotheses, and where they run out.

How do Hofstede and GLOBE differ?

They differ in source data, number of dimensions and whether they ask what people want. Hofstede's scores on the Culture Factor Group page are one score per dimension, drawn from scientific journals and the group's own projects; GLOBE was a ten-year programme by 170 researchers, who surveyed more than 17,000 managers in telecommunications, food and banking across 62 societies (Javidan et al., 2006, p. 69).

The two camps have argued in public. The abstract of the GLOBE team's reply says there is no basis for Hofstede's criticism that GLOBE's value measures are too abstract, and calls his reanalysis of the GLOBE data inappropriate (Javidan et al., 2006). I read only the abstract, so I do not referee. For a change designer the useful part is that leading researchers disagree on how to measure national culture, so both sets of scores are estimates.

Which numbers exist for Vietnam, and how far can they be trusted?

The table sets the page I opened on 3 November 2026 beside an archived copy of 1 July 2022 of the older address.

Four things deserve suspicion.

First, two numbers moved by 10 points between the two archived copies, and a copy of 31 March 2024 already carries the new figures. That copy and the live page both give the latest update as 16 October 2023, for individualism and long-term orientation, which matches the two numbers that changed. The page does not say whether Vietnam's figures moved because of new data or a new calculation.

Second, the page names no source specific to Vietnam. Its general source note at the foot says scores are updated when published in scientific journals, and that countries not covered there were added through the group's own studies or commercial projects; it lists references including Minkov and Kaasa (2022) and the 2010 edition of Software of the Mind. I do not know which group Vietnam's numbers belong to.

Third, the country list on the GLOBE project page has no Vietnam. Li and colleagues assigned GLOBE scores by the country where each study was run, and treated countries absent from GLOBE as missing (Li et al., 2021).

Fourth, anyone wanting GLOBE scores for Vietnam would have to borrow a country or cluster, and that is no better. GLOBE's Confucian Asia cluster holds China, Hong Kong, Japan, Singapore, South Korea and Taiwan. On the practice scores (what is), future orientation runs from 3.75 in China to 5.07 in Singapore, and in-group collectivism from 4.63 in Japan to 5.80 in China, both on a 7-point scale (Javidan et al., 2006, pp. 86–87). A societal cluster cannot tell you where one member sits inside it.

What should each score become when you design a change?

Treat each score as a hypothesis tied to one observation and one design decision, then test it in your own firm before you act. I leave out achievement motivation and indulgence because I cannot tie them to any decision about change. The first column paraphrases the Culture Factor Group page; the other two are mine.

By my estimate this costs about a week of one-to-one conversations, and in return you learn which scores hold for your own people. In a team of a handful you can talk to everyone anyway, so a national score adds nothing.

A hypothetical example: a 45-person distributor changes how it takes orders

Entirely hypothetical: the firm, the people and the numbers are my inventions, and they describe no real business. An electrical-equipment distributor in Bien Hoa has 45 staff, including 12 salespeople and 6 warehouse workers. Ms Van, the managing director, wants every order from 1 February entered in software instead of on paper slips and Zalo messages. Following the table, she announces the change at an all-staff meeting, then sees 8 people alone (6 sales, 2 warehouse) for 15 minutes each and asks what worries them, whom they would ask before trying it, and where the last change got stuck.

Invented result: all 8 were silent in the meeting, as the power distance hypothesis predicted. But 5 had a specific worry, and 3 of those feared the software would record who closed each order, so commission would be scrutinised. Two salespeople under 30 wanted to start the following week. One long-serving storekeeper said he would comply if the warehouse head went first.

The national score predicted the silence and missed what the worries were about. The commission worry falls under the tenth reason on Kanter's list, a real threat; according to Kanter's framework in the site's library, the response is to be honest, quick and fair (see Kanter's ten reasons people resist change). Ms Van takes that to mean stating the commission rule before launch. She changes three things: she writes the commission rule down, she runs a four-week pilot with the two young salespeople and the warehouse head, and she moves the cut-over date to 1 March. The price is a month's delay.

When should you drop national scores?

When you already have data on your own people, because that data answers the question better than a country average. The Culture Factor Group page itself says that differences among the values of individuals within one country tend to be bigger than the value differences between countries, and that country scores are generalisations that mean something only relative to other countries. It is the flaw of averages at the scale of a nation.

On participative leadership the ten GLOBE clusters score between 4.97 and 5.86 out of 7, and the GLOBE team writes that it is generally reported to contribute to outstanding leadership, though countries and clusters differ meaningfully (Javidan et al., 2006, pp. 73–74). Li and colleagues tested 56 pairings of leadership style and GLOBE score: 10 were statistically significant, most ran against their own hypotheses, and after a Bonferroni correction (p = 0.00089) none remained significant (Li et al., 2021). The post on cultural intelligence covers that study in more detail.

Your firm's own culture is a separate matter, found in Schein's three levels. A national score says nothing about it.

Going deeper: what did I read, and what did I not?

• Javidan et al. (2006) in Academy of Management Perspectives: full text, via a PDF on a personal website rather than the publisher's; page numbers follow that copy. The same authors' article in Journal of International Business Studies is paywalled and I read only the abstract. I have not read Hofstede's reply.
• Li et al. (2021): full text, via the Utrecht University repository.
• Culture Factor Group: the Vietnam entry, the general source note at the foot of the page and two archived copies; I found no methodology note specific to Vietnam. I have not read the works the page lists. The page allows non-commercial use of its data and I quote six numbers.
• GLOBE project page: the country list. I have not read House et al. (2004), so I cannot rule out Vietnamese data held elsewhere.

Two web pages outside the site's library, used as data sources rather than scholarly sources, in Harvard style:

• The Culture Factor Group (2026) Country comparison tool (formerly Hofstede Insights). Available at: https://www.theculturefactor.com/country-comparison-tool (Accessed: 3 November 2026).
• GLOBE Project (2026) 2004, 2007 studies. Available at: https://globeproject.com/study_2004_2007 (Accessed: 3 November 2026).