HRM304 · Session 2 of 11

Cultural frameworks,
and their limits

Lukas Wallrich · Birkbeck, University of London
Visiting Lecturer · SWUFE School of International Business · May 2026

Today, in three parts

Chunk 1测量文化

The frameworks. Hofstede, GLOBE, tightness–looseness. Why people tried to measure culture, and what they found.

Chunk 2应用

Applying them. A prediction game: can the numbers tell you what's going on in the room?

Chunk 3局限

What frameworks miss. Critiques, the institutional lens, and a Brazil case to test what we've learned.

Before we begin

Reflection 1反思

Have you started? You only need to submit three at the end — but writing about every session gives you choice and helps you learn.

Presentations小组报告

Groups and case topics are distributed at the end of today's session.

The waterline水线

Remember the iceberg? When we discuss HR practices, ask: above or below?

Chunk 1 of 3 · ~45 min

Can we measure culture?

Three attempts — and what each one was actually trying to do.

Why try to measure it at all?

Comparison. If two countries differ, by how much on what?

Prediction. Will this HR practice transfer — or will it break?

Diagnosis. When something fails abroad, which difference might explain it?

Every measure is a simplification. Always consider what it misses.

Cultural dimension

文化维度

A continuous axis along which national groups are claimed to differ on average — e.g., how much hierarchy is expected, how comfortable people are with ambiguity. A way of turning a fuzzy concept into a number you can compare across borders.

Hofstede's six dimensions

Power distance: a hierarchical pyramid of figures with a crowned figure at the top, contrasted with a flat row of equal figures.

Individualism vs collectivism: a single figure standing alone next to a networked group of figures connected by lines.

Masculinity vs femininity: a trophy on one side, two embracing figures inside a heart on the other.

Achievement vs Care prev. Masculinity vs Femininity

Uncertainty avoidance: a checklist contrasted with a winding path.

Long-term vs short-term orientation: a mature tree contrasted with a flower in bloom.

Indulgence vs restraint: an ice-cream cone and a music note contrasted with a padlock.

Each panel reveals on a click — pace the explanation, don't dump all six at once. Brief lines:
• Power distance — how much hierarchy people accept as legitimate. High: subordinates expect to be told. Low: bosses are challenged.
• Individualism vs collectivism — does identity sit in the person or the group? Affects how performance is attributed.
• Masculinity vs femininity — competition and achievement, vs care and quality-of-life. Most-criticised label today; newer versions use "achievement vs care."
• Uncertainty avoidance — tolerance for ambiguity. High: detailed rules and processes. Low: improvise.
• Long- vs short-term orientation — patience for delayed reward and willingness to adapt traditions.
• Indulgence vs restraint — how freely a society permits enjoyment and gratification.
Origin: Hofstede surveyed IBM employees in 50+ countries in the 1970s–80s, scored each on a 0–100 scale per dimension. The fifth and sixth dimensions were added later. Flag and move on: the masculinity label is the most contested today.

What Hofstede says about us

Power distance

UK 35

CN 80

Individualism

CN 43

UK 89

Achievement
(prev. masculinity)

CN · UK 66

Uncertainty avoidance

CN 30

UK 35

Long-term orientation

UK 51

CN 77

Indulgence

CN 24

UK 69

Scores from theculturefactor.com, retrieved May 2026 · scale 0–100.

Don't walk through all six — let students read. Lead with one: "Individualism — UK 89, China 43. The promotion conversation looks different on each side. In the UK, performance is attributed to the individual contributor. In China, far more to the unit. Same word, different referent." Pause. Then maybe one more: "Indulgence — UK 69, China 24. That's a workplace question too — about rest, holidays, after-hours expectations, what 'work-life balance' even means." Note the achievement/masculinity score: identical for both at 66. Useful to flag — high scores can come from very different underlying patterns. Source caveat: hofstede-insights.com rebranded to theculturefactor.com; some scores have been re-estimated in recent years (e.g., China individualism moved from 20 to 43) — that itself is a useful data point on framework fragility.

GLOBE — what counts as good leadership?

Anglo

Charismatic

Participative

Humane

Team-oriented

Autonomous

Self-protective

Germanic Europe

Autonomous

Charismatic

Participative

Humane

Team-oriented

Self-protective

Confucian Asia

Self-protective

Team-oriented

Humane

Charismatic

Autonomous

Participative

Southern Asia

Self-protective

Charismatic

Humane

Team-oriented

Autonomous

Participative

Sub-Saharan Africa

Humane

Charismatic

Team-oriented

Participative

Self-protective

Autonomous

Charismatic — inspires through vision and values.

Team-oriented — builds cohesion and shared purpose.

Participative — involves followers in decisions.

Humane — supportive, generous, considerate.

Autonomous — independent, self-reliant decision-maker.

Self-protective — status-conscious, face-saving; shields leader and in-group.

What stands out in each cluster (House et al., Culture, Leadership and Organizations, 2004). 5 of 10 clusters shown. Next slide shows the underlying numeric scores.

First-pass GLOBE slide — emphasises what is distinctive about each cluster's leadership preferences, not within-cluster absolute rank. (Charismatic is the top-scoring style in every cluster in absolute terms; that pattern is shown on the next slide.) GLOBE: 62-society project, 17,300 managers, 951 organisations, led by Robert House. Asked two questions per dimension — practices ("how it is") and values ("how it should be"). This slide is about preferred leadership (the CLT — culturally endorsed implicit leadership theory). Reveal one cluster at a time.

Highlights to draw out, cluster by cluster:
• Anglo — Charismatic + Participative on top, Self-protective at the bottom. Reads like the textbook "modern leader" — inspirational and inclusive. Most HR textbooks were written from inside this cluster.
• Germanic Europe — Autonomous on top. Counter-intuitive: students often imagine Germans as consensus-driven. GLOBE says Germanic managers most value an independent, self-reliant decision-maker.
• Confucian Asia — Self-protective on top, Participative at the bottom. This is the slide's most surprising finding. Face-saving, status-conscious behaviour is positively valued; subordinates rating bosses, flat hierarchies, town-hall challenges to authority do not read as "good leadership." 360-degree feedback imports fail here for this reason. Caveat to flag aloud: this is "most distinctive," not "highest absolute score" — see next slide.
• Southern Asia — Self-protective on top too, then Charismatic and Humane. The leader is expected to shield the in-group and inspire.
• Sub-Saharan Africa — Humane on top. Generosity, pastoral care, looking after the team's wellbeing is the core of legitimate authority.

HR implication to flag: succession plans, talent-spotting criteria, and 360 feedback systems all encode an implicit leadership theory. When global firms export "good leader = participative + inspirational" they're exporting an Anglo template that doesn't travel.

GLOBE — the absolute scores

Bar = cluster score on the 1–7 endorsement scale ·= 10-cluster average

Anglo

Germanic Europe

Confucian Asia

Southern Asia

Sub-Saharan Africa

Charismatic

6.05

5.93

5.63

5.97

5.79

Team-oriented

5.74

5.62

5.61

5.86

5.70

Participative

5.73

5.86

4.99

5.06

5.31

Humane-oriented

5.08

4.71

5.04

5.38

5.16

Autonomous

3.82

4.16

4.04

3.99

3.63

Self-protective

3.08

3.03

3.72

3.83

3.55

Charismatic — inspires through vision and values.

Team-oriented — builds cohesion and shared purpose.

Participative — involves followers in decisions.

Humane — supportive, generous, considerate.

Autonomous — independent, self-reliant decision-maker.

Self-protective — status-conscious, face-saving; shields leader and in-group.

Cluster means from House et al., Culture, Leadership and Organizations (2004), Table 21.5. 5 of 10 clusters shown.

GLOBE: 62-society project, 17,300 managers, 951 organisations, led by Robert House. Asked two questions per dimension — practices ("how it is") and values ("how it should be"). This slide is about preferred leadership (the CLT — culturally endorsed implicit leadership theory). Reveal one row (one leadership style) at a time so students can scan across all five clusters before moving on.

How to read the bars: each cell shows the cluster's absolute endorsement score on the 1–7 scale. The vertical tick on every bar marks the 10-cluster average for that style — so a bar that extends past the tick is above-average for the world; one that falls short is below-average. This lets the same chart tell two stories: the columns show each cluster's profile (read top to bottom); the rows show how the world spreads on each style (read left to right).

What to draw out, row by row:
• Charismatic — universally endorsed (all five clusters near or above the tick). Confucian Asia is the lowest of the five but still very high in absolute terms. The "inspirational leader" idea travels well.
• Team-oriented — also universally endorsed, narrow spread. Effectively a universal too.
• Participative — here the spread opens. Germanic Europe and Anglo well above the tick; Confucian Asia and Southern Asia clearly below. This is where the "good leader = inclusive decision-maker" Anglo assumption stops travelling.
• Humane-oriented — Southern Asia and Sub-Saharan Africa above average; Germanic Europe below. Confucian Asia is around the average — slightly above.
• Autonomous — narrowest spread. Germanic Europe edges above the tick (the often-quoted "Germanic autonomy" finding); the others all sit close to average.
• Self-protective — this is the slide's headline. In absolute terms it's every cluster's lowest-scoring style. But Confucian Asia, Southern Asia, and Sub-Saharan Africa sit above the world average, while Anglo and Germanic sit well below. So self-protective leadership (face-saving, status-conscious, shielding in-group) is genuinely positively endorsed in Confucian Asia — not the top-scoring style for them, but distinctively more endorsed than in Anglo culture. 360-degree feedback imports run into exactly this gap.

Headline insight: rows 1–2 are universals, rows 5–6 are the most culturally contingent. The Anglo "modern leader" template — charismatic + participative — is partly universal (charismatic travels) and partly Anglo-specific (participative doesn't travel as well, especially eastward).

HR implication to flag: succession plans, talent-spotting criteria, and 360 feedback systems all encode an implicit leadership theory. When global firms export "good leader = participative + inspirational" they're exporting an Anglo template that partly travels — and they need to know which parts don't.

Gelfand — how much room is there to deviate?

Tight cultures enforce social norms strongly; deviance is sanctioned quickly. (e.g., Japan, Singapore, Pakistan)

Loose cultures tolerate a wider band of behaviour. (e.g., Netherlands, Brazil, New Zealand)

China scores relatively tight — but the data is from before 2015. Is that changing in younger urban cohorts?

Tightness vs looseness illustration in line-art style: top half shows figures aligned in a strict grid labelled TIGHT, bottom half shows figures scattered freely labelled LOOSE, with terracotta accents.

Meyer — the Culture Map

Meyer's Culture Map chart. Four horizontal scales with terracotta dots showing country positions. Communicating: US/NL near low-context, Japan/China near high-context. Leading: Denmark/NL near egalitarian, China/Korea near hierarchical. Disagreeing: France/Germany near confrontational, Japan/China near avoids-confrontation. Trusting: US/NL near task-based, China/Brazil near relationship-based.

Eight behavioural scales — practitioner-built, four shown. Positions are illustrative, not a measured dataset.

Erin Meyer, INSEAD professor. Book: The Culture Map (2014). Not academia-first like Hofstede or GLOBE — written for managers, drawn from her consulting work. Eight scales total; the four above are the ones HR people meet most often. Other four: Persuading (principles-first ↔ applications-first), Deciding (consensual ↔ top-down), Scheduling (linear-time ↔ flexible-time), Evaluating (direct negative feedback ↔ indirect).

Why this matters as a fourth framework:
• Comparison, not absolutes. Meyer's point: don't ask "is China hierarchical?" — ask "is China hierarchical relative to whom?" A US manager experiences China as hierarchical; a Korean manager experiences it as middling. The scale only makes sense in pairs.
• Behaviour, not values. Hofstede asks what people value; Meyer asks how they behave at work. More directly usable in an HR meeting.
• Decouples dimensions Hofstede conflates. "Leading" (who decides) and "Deciding" (how — consensus vs top-down) are separate. Japan is hierarchical on Leading but consensus-driven on Deciding. Hofstede's power-distance hides that.

Where to push back: the evidence base is anecdotal — drawn from her clients and training sessions, not a replicable dataset. Strong on resonance, weak on rigour. Use it like a phrasebook, not a thermometer.

Verbal hook: "If Hofstede is the cartographer, Meyer is the travel guide."

Four frameworks, four vibes

Hofstede

霍夫斯泰德

Six numbers per country. Comparable, blunt, ageing.

GLOBE

全球研究

Focused on leadership. Practices vs values. Directly HR-applicable.

Tightness–looseness

紧–松

How much variation a society tolerates. Newer, narrower.

Culture Map

文化地图

Eight behavioural scales, read in pairs. Practitioner-built.

Chunk 2 of 3 · ~45 min

The Hofstede prediction game

Can the scores tell you what's happening in the room?

What you'll have in front of you

A scenario

One paragraph describing a workplace miscommunication or HR failure in a specific country. Every group gets a different one.

A score sheet

Hofstede scores for the country in the scenario, side-by-side with China. Six dimensions, scale 0–100.

Group Task 小组任务 20:00 · 4 per group

Your task

Use the Hofstede scores to predict what's happening in the scenario — and what the scores can't explain.

What to do

Read the scenario and the score sheet. Underline anything that surprises you.
Write two short paragraphs: (a) what the scores predict, (b) what they miss.
Choose a spokesperson for a 90-second share-back.

Report-back 反馈 ~90s per group

What did the scores get right? What did they miss?

How we'll do it

Spokesperson: 90 seconds. Country + one prediction + one limit.
I'll cluster recurring limits on the board.
Other groups: build on patterns you also saw.

Chunk 3 of 3 · ~45 min

What the frameworks miss

And why culture is only half the story.

Ecological fallacy

生态谬误

Treating a group average as if it applied to every individual in the group.

Saying "Chinese employees are collectivist" because the country score is 43 — when an individual in front of you might well be at 80.

Four limits of frameworks

One number per country

Within-country variation is huge — regional, generational, individual.

Frozen in time

1970s IBM data, patched but not refreshed. Hofstede argued culture is stable — generational data says otherwise.

A Western lens

Built from Anglo/Dutch academia. What was not asked?

Culture ≠ institutions

Many "cultural" differences are really legal or political.

Culture is one lens. Here are three more.

Law法律

Mandatory works councils in Germany. At-will employment in the US. Dual-track unions in China. HR cannot opt out of legal frames.

Labour markets劳动力市场

Tight or loose? Surplus or shortage? Do workers get to choose? Shapes what HR can demand and what it must offer.

Political economy政治经济

Liberal vs coordinated vs state-influenced capitalism. Influences whose interests HR is permitted to serve.

Walmart in Germany — a $1bn lesson

95 stores acquired 1997–98 (Wertkauf, Interspar) · Exited 2006 · Sold to Metro at an estimated $1bn loss

The surface — cultural irritants

Mandatory morning chants and warm-up exercises — staff refused, found them infantilising.

Greeters smiling at every customer — read as insincere; sometimes mistaken for flirtation.

"10-foot rule" — greet any customer within 10 feet. German shoppers found it intrusive.

US-based head ran meetings in English; refused to learn German.

The cause — institutional collisions

Works councils (Betriebsrat) sidelined — Mitbestimmung makes co-determination mandatory.

"No-dating" ethics code struck down by Wuppertal labour court (2005).

Predatory pricing to undercut Aldi/Lidl — blocked by the Bundeskartellamt.

Slow, reluctant recognition of the Ver.di union; repeated strikes followed.

Culture names the symptoms. Institutions name the cause.

Worked case after the institutional-lens slide. Spend ~5 min telling it; then use it as the reference point when the Brazil consultant task starts.

Quick narrative arc to tell from the slide:
1. 1997–98: Walmart enters Germany by buying Wertkauf (21 hypermarkets) and Spar/Interspar (74 stores). Ambition: replicate the US "always low prices" model in Europe's biggest market.
2. The visible failures — the left column. These are what made the press. Easy to laugh at, easy to over-diagnose as "Germans don't smile / Americans are too friendly." That's the cultural reading, and it's not wrong — just superficial.
3. The killers — the right column. These are the structural reasons Walmart could not operate at scale: Mitbestimmung (co-determination, Betriebsverfassungsgesetz 1972), labour-court protections, competition law (GWB §20 — predatory pricing), strong sector unions. None of these are "culture" — they're enforceable institutions.
4. 2006: sold to Metro AG; reported losses ~$1bn. Walmart's stated reason in its filings was an inability to achieve required scale economies. True, but only because the institutions blocked the moves that would have produced those economies.

Diagnostic move to model out loud: "If I'd been a cultural-frameworks consultant in 1997, I'd have predicted the chants would fail and the smiling would land oddly. I would not have predicted that the no-dating policy would end up in labour court, or that below-cost pricing would be illegal. That's the institutional layer — and that's what cost the billion dollars."

Sources (don't read on slide; reach for if challenged): Knorr & Arndt (2003) Why did Wal-Mart fail in Germany? Univ. Bremen IWIM working paper; Christopherson (2007) Barriers to ‘US style’ lean retailing, JEG; multiple HBS / case-study writeups.

Bridge into the Brazil task: "Same diagnostic move you just watched — now do it for Brazil. Don't stop at the cultural reading; ask which institutions are also doing the work."

Group Task 小组任务 15:00 · 4 per group

Scenario

A Chinese tech company expanding to São Paulo hires a consultant who tells them: "Adapt to high power distance and collectivism — Brazilians will respond like Chinese employees do." What's wrong with this advice?

What to do

Find as many distinct problems with this advice as you can. Use today's concepts.
Suggest one thing the company should do instead.

Hint: São Paulo Card 2 showed one version — what changes when the parent is Chinese?

Panel hides: labour law, generational shift, within-country variation.

Brazil vs China — the data you have

Power distance

BR 69

CN 80

Individualism

BR 36

CN 43

Achievement

BR 49

CN 66

Uncertainty avoid.

CN 30

BR 76

Long-term orient.

BR 28

CN 77

Indulgence

CN 24

BR 59

Meyer · Trusting both relationship-based

Meyer · Disagreeing BR more confrontational than CN

Gelfand BR loose · CN tight

Synthesis activity. The data panel is on-screen during the task — students should be reading, not recalling. The Hofstede comparison: PDI 69/80 and IDV 36/43 are the one pair-up the consultant has right (both on the same side); the other four dimensions diverge sharply.

What to listen for in share-back — five distinct critique types:
• Cherry-picking within Hofstede — the consultant used 2 of 6 dimensions; UAI (76 vs 30), LTO (28 vs 77), IVR (59 vs 24) all flip the playbook.
• Cross-framework divergence — Gelfand looseness and Meyer's Disagreeing scale say Brazilian workplaces won't absorb a Chinese-tightness playbook quietly.
• Institutions ≠ culture — CLT dismissal protections, mandatory 13th salary, vacation rules. A China-style high-churn / stack-rank playbook is partly illegal, not just culturally awkward.
• Ecological fallacy — São Paulo white-collar tech workers ≠ Brazil-on-average; advice for "Brazilians" elides huge within-country variation.
• Frozen-in-time / Western lens — 1970s IBM origins, framework not built to capture jeitinho or personalism.

The Card-2 hint is the cheapest path in for groups who freeze: Munich → São Paulo already taught the institutional + Meyer-Trusting moves. Ask: what's the same (CLT still bites, Brazilian relationship-trust still bites) and what's different when the parent is a Chinese tech firm (relationship-trust now overlaps; tightness/looseness becomes the bigger gap; hours/996 norms hit Brazilian labour law differently than German appraisal forms did)?

Verbal landing line for the wrap: "If you'd only read the consultant's two dimensions, you'd write the same playbook for Brazil as for China. The other four dimensions tell you why that playbook will misfire — and the institutional layer tells you why the misfire will be expensive."

Report-back 反馈 30s per group

One problem with the consultant's advice — your sharpest one.

Group presentations — your groups and your case

How groups formed

Randomly assigned, 5–6 per group. List will be online by end of day.

Your case

Each group has a multinational HR scenario. Case details will be online by end of day.

First milestone

A 1-paragraph problem statement, due before Session 5.

Next session

Communication & feedback

If performance means different things in different places — what happens when you sit down to tell someone how they're doing?

Cultural frameworks,and their limits

Today, in three parts

Before we begin

Can we measure culture?

Why try to measure it at all?

Hofstede's six dimensions

What Hofstede says about us

GLOBE — what counts as good leadership?

GLOBE — the absolute scores

Gelfand — how much room is there to deviate?

Meyer — the Culture Map

Four frameworks, four vibes

The Hofstede prediction game

What you'll have in front of you

What the frameworks miss

Four limits of frameworks

Culture is one lens. Here are three more.

Walmart in Germany — a $1bn lesson

Group presentations — your groups and your case

Communication & feedback

Cultural frameworks,
and their limits