Oxford Handbook Hrm

1. 1Position employee attitudes as the mediating link between HR practices and performance — not the endpoint. Model the cause-effect chain: HR policies → enacted by line managers → shape attitudes → drive productive behaviors → firm outcomes.
2. 2Decompose 'attitudes' into the specific constructs the passages name and measure them separately: job satisfaction, trust-in-management, attitudinal commitment. Treat these as intervening variables, not a single lump.
3. 3Distinguish attitude from behavior in your measurement: link attitudinal commitment to discretionary job behavior and behavioral commitment as downstream, separable outcomes.
4. 4Audit whether line managers actually enact the HR policies as designed — attitudes only improve if policies are 'effectively enacted by line managers.' Measure the gap between espoused policy and enacted practice.
5. 5Import constructs from organizational behavior — psychological contracting and social exchange — to explain why attitudes shift, rather than assuming practices mechanically produce satisfaction.
6. 6Fix your measurement design to establish causal direction: use genuinely longitudinal data measuring current HR practices against future performance, not cross-sectional studies that pair current practices with historical financial data.
7. 7Test for mutuality: assess whether outcomes are mutually satisfying to employer and worker, since mutually satisfying outcomes are 'more sustainable' over the long run. Don't measure only shareholder/management-sought outcomes.

Consider the advanced-manufacturing case the passages describe: direct labor cost is low, but workers have a major effect on how well the technology is utilized. Here the firm remunerates and trains workers very well and works to keep motivation high — a high-commitment model. The attitudinal logic is that better pay and training raise trust-in-management and attitudinal commitment, which convert into the discretionary behavior needed to run the technology well. In such sectors this attitude-shaping investment is a 'table stake': firms either adopt it or they won't survive. Note the passages' caution — employers do not adopt this across the board, because industry differences in productive technologies and customer attitudes determine which model is economically sustainable, and 'strategic tensions' between employer and employee persist no matter how appealing the model. So even a high-commitment system doesn't eliminate the underlying tension around labor motivation; it manages it within the firm's economic resources.

• If workers have a major effect on how well technology or service is delivered, then invest in attitude-shaping practices (pay, training, motivation) because cost effectiveness — not just cost — is the driver.
• If labor cost dominates and workers have little effect on utilization, then don't adopt high-commitment models across the board — industry technology and customer attitudes determine sustainability.
• If your HR-performance study is cross-sectional, then treat any association as suspect: it likely links current practices to past performance and can't support the claim that practices cause performance.
• If a policy exists on paper but line managers don't enact it, then expect no attitudinal improvement — enactment by line managers is a precondition for attitude effects.
• If you measure only outcomes sought by shareholders or managers, then you've missed mutuality — also test whether worker outcomes are satisfied, because that determines long-run sustainability.

• Treating employee attitudes as a single undifferentiated variable. The tell: your model has one 'satisfaction' box instead of separating job satisfaction, trust-in-management, and attitudinal commitment.
• Inferring causality from cross-sectional data. The tell: you asked about current HR practices while your performance figures are historical financial data — you've correlated practices with the past.
• Assuming policies produce attitudes automatically. The tell: no measurement of whether line managers actually enacted the practice as designed.
• Collapsing attitude into behavior. The tell: you can't distinguish attitudinal commitment from discretionary job behavior or behavioral commitment in your analysis.
• Measuring only management-sought outcomes and ignoring mutuality. The tell: no worker-side outcome in the model, so you can't judge whether the arrangement is sustainable long-term.
• Assuming a high-commitment model fits everywhere. The tell: you ignore industry differences in productive technologies and customer attitudes that determine which model is economically sustainable.

• ▸Worker impact on delivery: rate how much workers affect technology/service utilization
• ▸HR practice inventory: list the practices intended to shape attitudes (pay, training, voice)
• ▸Line-manager enactment: score gap between espoused policy and enacted practice
• ▸Attitude measures: separately record job satisfaction, trust-in-management, attitudinal commitment
• ▸Behavior measures: record discretionary job behavior and behavioral commitment separately
• ▸Data design: state whether attitude-performance data is cross-sectional or longitudinal
• ▸Causal direction check: confirm HR measures precede performance measures in time
• ▸Mutuality check: note whether worker outcomes (not just management's) are satisfied
• ▸Sustainability verdict: judge whether the model fits the industry's technology and customer attitudes

1. 1Start from the strategy, not the practices: specify the employee behaviors the strategy requires before you touch any HR practice. For an innovation strategy, that means behaviors like experimentation and discovery; for a discount-retail strategy, it means something else entirely (compare Nordstrom's workforce to Wal-Mart's).
2. 2Derive the skills that make those behaviors possible. Treat skills as necessary but not sufficient — scientific knowledge enables experimentation, but possessing it does not guarantee the behavior occurs. Map required skill for each required behavior.
3. 3Separate the three behavior-relevant components you are trying to move: the human capital pool (stock of knowledge, skills, motivation, behaviors), the flow of human capital through the firm (movement of people and knowledge, plus willingness to create, share, apply knowledge), and the dynamic renewal processes. Assign HR practices to each.
4. 4Design HR practices to both build the skill base AND elicit the relevant behavior — these are two distinct jobs. Ability-raising practices alone do not produce behavior; pair them with motivation and opportunity-to-contribute practices.
5. 5Route practices through line managers deliberately: HR policies only work well when effectively enacted by line managers. Build the enactment step in rather than assuming a policy on paper equals a policy in practice.
6. 6Instrument the mediating chain, not just the endpoints. Measure the intervening constructs — job satisfaction, trust-in-management, attitudinal commitment, discretionary job behavior, behavioral commitment — so you can see whether practices are actually shifting behavior before operational outcomes appear.
7. 7Trace the causal sequence to operational outcomes (quality, customer service, turnover) and only then to firm-level outcomes (revenues, profits). Do not attribute financial results directly to practices while skipping the behavior link.
8. 8Where effort is unobservable, choose an incentive mechanism to the behavior at issue: output-based pay to induce shirkers to quit and reward producers, wider internal wage spreads for tournament-style promotion effort, or steep seniority profiles with penalties to deter shirking.

Take the Nordstrom versus Wal-Mart contrast in the passages. Nordstrom pursues an upscale-retail strategy, so the required employee behaviors center on high-discretion customer service and initiative; Wal-Mart's discount strategy requires cost-disciplined, high-volume, standardized behaviors. Working backward, Nordstrom must first specify the service behaviors, then identify the skills that support them, then build a human capital pool and reward system that both develops those skills and elicits the discretionary behavior — because skills are necessary but not sufficient. Wal-Mart, by contrast, fits a different skill and behavior profile to its cost-focused strategy. The point the passages make is that the same list of HR practices cannot be optimal for both: fit means matching the skills and behaviors of employees to the firm's strategy. If you copied Nordstrom's practice bundle into Wal-Mart, you would be paying to elicit behaviors the strategy does not require. This is why 'best practice' counts of HR practices, while positively correlated with performance, still miss the strategy-specific behavior that fit demands.

• If the strategy requires a distinct behavior (e.g., experimentation), then specify that behavior first and derive skills from it — never start by selecting practices.
• If employees have the skills but the behavior is absent, then the gap is motivation/opportunity, not ability — add incentive and voice practices rather than more training.
• If effort is hard to observe and screening is prohibitively costly, then use output-based pay or steep seniority profiles with penalties instead of monitoring.
• If a policy exists on paper but line managers are not enacting it, then treat the line-manager enactment step as the failed link and fix that before adding practices.
• If two business units run different strategies, then expect different HR systems within the same firm — apply the human resource architecture (value and uniqueness dimensions) rather than one uniform bundle.

• Attributing firm financial performance directly to HR practices while skipping the human capital pool and behavior mediators. Tell: your model has practices on one side and profit on the other, with nothing measured in between.
• Believing more practices is always better. The evidence shows a correlation (roughly 20% better financial performance per standard deviation of practices), but Gerhart himself flags that effect size as 'so large as to perhaps not be credible' — treat headline numbers as suspect, not gospel.
• Single-respondent, single-survey measurement: one senior HR manager reports both which practices exist and how well the firm performs. Tell: common-source bias inflates the relationship and you cannot separate perception from reality.
• Copying a rival's practice bundle without matching behaviors to your own strategy. Tell: you adopt 'sophisticated HRM' wholesale and see no operational gain because the elicited behaviors don't fit your strategy.
• Assuming skills equal behavior. Tell: you invested heavily in training, competence rose, but discretionary behavior and operational outcomes did not move because motivation and opportunity were never addressed.

• ▸Business strategy: state the competitive stance in one line (e.g., innovative products vs. discount).
• ▸Required behaviors: list the specific employee behaviors that strategy demands (e.g., experimentation, discovery).
• ▸Enabling skills: for each behavior, name the skill/knowledge that makes it possible.
• ▸Component targeted: mark whether each practice moves the human capital pool, the flow, or renewal processes.
• ▸HR practices — ability: practices that build the skill base.
• ▸HR practices — motivation/opportunity: practices that elicit the behavior (incentives, voice, effort mechanism where effort is unobservable).
• ▸Line-manager enactment: name who enacts each practice and how enactment is verified.
• ▸Mediator measures: list attitudinal/behavioral constructs you will track (satisfaction, trust, commitment, discretionary behavior).
• ▸Outcome trace: operational outcome (quality, service, turnover) → firm-level outcome it should feed.

1. 1Position performance outcomes at the correct link in the causal chain: intended practices → actual practices → perceived practices → attitudinal outcomes → behavioral outcomes → performance outcomes. Do not skip from practices straight to performance.
2. 2Reject distal financial measures (shareholder value, published profitability) as your primary metric; they are too far removed from HRM influence and cannot explain why or how practices moved them.
3. 3Select proximal measures directly influenced by worker behavior: employee turnover, absence, scrap rates, sales per employee, productivity, and customer satisfaction.
4. 4Rank your proximal measures by expected strength of linkage to employee satisfaction/engagement — following Harter et al. (2002), expect customer satisfaction, loyalty, and turnover to correlate highest, then safety and productivity, with profitability lowest.
5. 5Adjust every measure for sectoral variation: nursing and teaching show higher sickness rates; retail firms and call centers show higher turnover, so benchmark within sector, not across.
6. 6Structure comparisons across units running identical HR practices (e.g. bank branches) to isolate the practice effect, as Gelade and Ivery (2003) and Bartel (2004) did in retail banking.
7. 7Use longitudinal, not cross-sectional, data — cross-sectional studies pick up historical financial data while asking about current practices, so they cannot establish that causality runs from practice to outcome.
8. 8Combine employee surveys, customer surveys, and operational measures into one linked dataset so you can trace attitudes → behavior → performance rather than asserting the connection.

A large retail bank wants to know whether its people-management practices raise performance. Instead of using group-level profitability, it exploits the fact that hundreds of branches run identical HR practices, following the approach of Gelade and Ivery (2003) and Bartel (2004). It surveys employees for satisfaction/engagement, surveys customers for satisfaction and loyalty, and pulls operational data on turnover, absence, and sales per employee. Consistent with Harter et al. (2002), the strongest correlations with engagement appear for customer satisfaction, loyalty, and turnover, with productivity weaker and profitability weakest. Because the branches share practices and are compared cross-sectionally against each other rather than over time, the bank treats the branch as the unit and adjusts for local labor-market differences in turnover. The resulting model tells managers something actionable — which behaviors move customer satisfaction — rather than the uninformative 'more HR practices the better' of the Huselid curve.

• If a measure is financial and published (profitability, shareholder value), then treat it as distal and demote it — you cannot explain how HR influenced it.
• If a measure flows directly from worker behavior (turnover, absence, scrap, customer satisfaction), then use it as your proximal outcome.
• If your data are cross-sectional, then do not claim causality from practice to performance — you may only be capturing past performance predicting current practice.
• If you are comparing units, then require identical HR practices and adjust for sector before attributing differences to people management.
• If you must rank which outcome to prioritize, then follow the Harter ordering: customer satisfaction and loyalty and turnover first, productivity and profitability last.

• Jumping straight from HR practices to performance and skipping attitudinal and behavioral mediators — the tell is a model with no measured 'why' or 'how' between practice and result.
• Using financial performance as the outcome — the tell is being unable to give any convincing account of how a practice produced the number, leaving only 'more practices the better.'
• Running a cross-sectional design and asserting causality — the tell is asking about current practices while the performance data are actually historical.
• Ignoring sector adjustment — the tell is flagging nursing/teaching as 'high absence' or call centers as 'high turnover' as if it were a management failure rather than a sector baseline.
• Assuming firms structure HR to maximize short-term financial outcomes — the tell is measuring only shareholder value and missing outcomes that matter to employees, so results are unsustainable and one-sided.

• ▸Outcome measure: name the specific proximal metric (e.g. turnover, scrap rate, customer satisfaction).
• ▸Behavioral driver: state which employee behavior directly produces it.
• ▸Distal/proximal check: mark whether the measure is close to or far from worker influence.
• ▸Harter rank tier: place it in expected correlation order (high: satisfaction/loyalty/turnover; mid: safety/productivity; low: profitability).
• ▸Sector adjustment: note the sector baseline you must correct for (e.g. nursing sickness, call-center turnover).
• ▸Comparison unit: identify the unit (branch, team) and confirm identical HR practices across it.
• ▸Data design: state whether data are longitudinal or cross-sectional and the causality limit.
• ▸Data sources combined: list which of employee survey, customer survey, operational, and financial data you are linking.

1. 1Specify your effectiveness construct as a set, not a single number. Boxall and Purcell define superior HRM performance as 'an outstanding mix of outcomes' across three areas: productivity (short-term performance), flexibility (long-term viability, agility, sustained competitive advantage), and social legitimacy (meeting societal expectations and regulations that vary by country). Name your target explicitly across all three.
2. 2Choose financial indicators deliberately, knowing they are distal — too far removed from HR influence to explain how or why practices matter. If you use profitability, return on assets, or market value, use Tobin's q (the ratio of a firm's market value to the replacement cost of its assets) as Huselid did, and treat it as a downstream endpoint rather than the direct dependent variable.
3. 3Build a path model that specifies both proximate and distal dependent variables. Follow the revised HR causal chain: intended practices → actual practices → perceived practices → attitudinal outcomes → behavioral outcomes → performance outcomes (proximal or distal). Do not jump straight from HR practices to financial results.
4. 4Insert mediators between practices and financial outcomes. Huselid tested turnover and productivity as more proximate endogenous variables mediating the impact of high-performance work practices on financial performance; only the productivity → Tobin's q path held. Model these paths rather than assuming a direct effect.
5. 5Adjust every measure for sector. Published corporate accounts cover different business units and countries with different HR systems, driven by different regulatory or institutional regimes. Proximal measures like turnover and absenteeism also need adjustment for sectoral variants before comparison.
6. 6Control for other causal influences and set realistic expectations for variance explained. Firm performance is a function of many variables inside and outside the firm; Hitt et al. found human capital explained 3.6% of the variance in firm performance and treated that as significant. Do not read a small effect as a failed one.
7. 7Use social exchange theory to explain the downstream link. Attitudinal outcomes (satisfaction, willingness to cooperate) drive behavioral outcomes (OCB, attendance, retention), which feed performance. Gerhart concludes positive workforce attitudes create value; measure attitudes as the mechanism connecting practices to effectiveness.

Huselid (1995) is the only recruitment study that specified a full path model instead of a direct practice-to-money link. He expected turnover and productivity, as more proximate endogenous variables, to mediate the impact of high-performance work practices — including a factor containing recruitment intensity — on financial performance. When tested, the HR variable holding recruitment intensity was not related to one mediator and one dependent variable, so only a single mediation effect survived: through productivity as mediator to Tobin's q, the market-value-to-replacement-cost ratio. This demonstrates both the discipline and the fragility of distal effectiveness modeling: most of the hypothesized paths did not hold, and the surviving one ran through a proximal mediator rather than directly to the financial endpoint. The lesson is that a positive association at the aggregate level dissolves without a specified mechanism, which is why Purcell and Kinnie call financial data 'remarkably uninformative from a practitioner point of view beyond the knowledge that the more HR practices the better' — the so-called Huselid curve.

• If your only outcome is short-term shareholder value, then you have mis-specified the construct — add long-term viability and social legitimacy, because firms do not necessarily structure HR systems to maximize financial outcomes.
• If you cannot explain why or how an HR practice influences a financial measure, then move to proximal measures (turnover, absence, scrap rates, sales per employee, customer satisfaction) that are directly influenced by worker behavior.
• If you must rank proximal outcomes by strength of link to attitudes, then expect Harter et al.'s ordering: customer satisfaction, loyalty, and turnover highest; then safety and productivity; profitability lowest.
• If you compare organizations across sectors or countries, then adjust the measures — published accounts and turnover/absenteeism figures are confounded by different regulatory and institutional regimes.
• If the environment demands flexibility and agility, then treat long-term viability as possibly the dominant outcome requirement, even over short-term performance.

• Treating financial performance as the whole of effectiveness. The tell: your model has no legitimacy or viability outcome and assumes firms maximize short-term shareholder value.
• Running a direct practices-to-financials regression with no mediators. The tell: you report a 'more HR is better' correlation (the Huselid curve) but cannot state the causal path.
• Adopting the resource-based view uncritically as your effectiveness theory. The tell: your propositions are true by definition (tautological) and cannot be disconfirmed empirically — a criticism the RBV's own field has raised.
• Combining recruitment (or any single practice) with other variables into one latent construct. The tell: you cannot separate the effect of the practice you care about because the factor structure buries it.
• Dismissing a small effect size as failure. The tell: you discard a 3.6% variance-explained result without accounting for the many other causal influences on firm performance.

• ▸Productivity target: the short-term performance outcome you seek (name the metric)
• ▸Flexibility/viability target: the long-term agility or sustained competitive advantage requirement
• ▸Social legitimacy target: the societal expectation or country-specific regulation to be met
• ▸Distal financial indicator: chosen metric (e.g., Tobin's q, ROA, market value) marked as endpoint
• ▸Proximate mediators: worker-influenced measures (turnover, productivity, absence, customer satisfaction)
• ▸Attitudinal link: the satisfaction/cooperation measure connecting practices to behavior (social exchange mechanism)
• ▸Sector adjustment: the confounds (business unit, country, regulatory regime) you will control for
• ▸Path statement: written intended→actual→perceived→attitudinal→behavioral→performance chain
• ▸Expected variance: realistic effect-size range given other inside/outside causal influences

1. 1Separate practice from process. First map the system of HR practices you have in place; then, following Bowen and Ostroff (2004), audit the supporting processes that determine how employees actually perceive those practices. The perception layer is where climate lives.
2. 2Distinguish psychological climate from collective climate. Individuals form psychological climates from their own experiences; your job is to build consistency across people so those individual perceptions converge into a shared collective climate, which is what strengthens the HR system.
3. 3Instrument the line-manager layer. Because much of the psychological contract is negotiated locally between employee and line manager (LMX-style social exchange), treat line managers as agents of the organization and confirm they understand the obligations they create through informal arrangements — otherwise climate drifts outside HR's control.
4. 4Engineer perceived organizational support. Use consistent, visible signals (Eisenberger et al. 1986) so employees read the organization as supporting them; consistency across time and across managers is the mechanism that raises HR-system strength.
5. 5Contextualize before importing practices. Analytical HRM is sceptical of universal 'best practices' (contra Pfeffer's lists). Check moderators — competitive choices, technology, workforce characteristics, labor markets, national culture and law — before assuming a practice that built strong culture elsewhere will do so here.
6. 6Locate culture inside the goal set. Read culture against both economic objectives (labor productivity, cost-effective labor management, viability) and socio-political objectives, since employers adapt to the socio-political climate of work inside and outside the firm.
7. 7Treat culture as dynamic capability. Assess whether your people-management fosters a culture in which strategies get conceived and reworked over time, or, if handled badly, hinders that dynamic capability — because it is executed strategy that counts (Kaplan and Norton).

Consider the Bowen and Ostroff (2004) reasoning applied to a firm rolling out a new appraisal practice. HR designs the practice, but employees experience it only through their line managers, who make informal, local arrangements that shape each individual's psychological climate. In one team the manager applies the practice consistently and reciprocates effort with support, so employees perceive strong organizational support (Eisenberger et al. 1986); in another the manager improvises, and perceptions fragment. Because the second manager acts as an uninstructed agent of the organization, the intended message of the HR system is diluted and no shared meaning forms. Gelade and Ivery (2003) support the view that this climate layer mediates whether the practice reaches performance at all. The corrective is not more practices but consistency of process — turning scattered psychological climates into a collective climate that makes the HR system 'strong.'

• If a practice worked in another country or firm but the moderators differ (capital patience, culture, law, labor market), then treat it as a hypothesis, not a prescription, and test before scaling.
• If key elements of the psychological contract are being set locally by line managers, then instrument and train those managers as organizational agents before adding any new HR practice.
• If individual perceptions of climate are fragmented, then invest in consistency and visible organizational support to convert psychological climates into a collective climate rather than adding more programs.
• If leadership starts with a favorite practice (e.g. 360-degree appraisal) and cannot state the overarching HRM principle it serves, then stop and define the principle and goal first.
• If you cannot show the cause–effect chain from practice to employee attitudes to behavior to performance, then treat any practice–performance correlation as unproven and suspect confounding.

• Measuring practices and firm performance and claiming a causal link — the tell is an association found but never traced through the mediating attitudinal and behavioral chain (Wall and Wood 2005; Wright et al. 2005).
• De-contextualized 'best practice' transplanting; the tell is a prescription list applied identically across contexts with no mention of moderators (the Legge 1978 / Marchington and Grugulis critique).
• Leaving line managers uninstructed so informal local deals drift out of HR's control; the tell is inconsistent employee experience of the same formal policy across teams.
• Starting from favorite techniques rather than principles; the tell is a program launched with no articulated overarching HRM goal it serves.
• Ethnocentric, top-tier sampling — building your culture playbook from a few elite firms in a few countries; the tell is prescriptions that ignore the wide frequency distribution of firms by breadth and depth of HRM.

• ▸Target practice: name the HR practice whose perceived impact you are auditing.
• ▸Formal signal: state what the practice is intended to communicate to employees.
• ▸Line-manager enactment: describe how managers actually deliver/interpret it locally.
• ▸Psychological climate spread: note where individual perceptions diverge across teams.
• ▸Collective climate check: is there shared meaning, or fragmentation? (yes/no + evidence).
• ▸Perceived organizational support cues: list the consistent signals reinforcing support.
• ▸Moderators present: capital patience, culture, law, labor market, workforce traits at play.
• ▸Cause-effect chain: map practice → attitudes → behavior → performance for this case.
• ▸Corrective action: consistency/training step to convert psychological into collective climate.

1. 1Start by choosing your unit of analysis: don't analyze the individual firm in isolation. Map the sector and occupational context (manufacturing, service, knowledge-intensive, public) because work system archetypes and their constraints differ sharply across them.
2. 2Classify the work using a taxonomy of work content characteristics (as in Table 10.1) to identify which work system archetype applies, then match a job design strategy to it (per Table 10.2) rather than defaulting to one universal model.
3. 3Make explicit make/buy (insource/outsource) decisions for each job category — but base them on the availability of ready-trained labor in the external market, not only on the strategic uniqueness of the skill.
4. 4For any category treated as 'commodity' or externally contracted, do not stop at rule-compliance: build daily mechanisms to keep incumbent staff motivated, working in the client's interest, and cooperating with other parts of the production chain.
5. 5Analyze interactions between job categories, not just the value of each skill in isolation — operational efficiency depends on the coordination between roles across the production chain.
6. 6Test your work structure against the institutional environment: for transferable-skill training, confirm that state, employer associations, and unions supply information circulation, deliberation, monitoring, and sanctioning — absent these, firms rationally underinvest for fear of poaching.
7. 7In service settings, integrate marketing, operations, and HR deliberately, and name the tension: operations pushes efficiency and cost reduction while marketing pushes customer satisfaction, creating conflicting pressures on how work is designed.
8. 8Seek a strategic balance between economic objectives (competitiveness, efficiency) and justice/legitimacy objectives, accepting that the employment transaction is incomplete a priori and outcomes remain indeterminate.

Consider the training decision analyzed through the varieties of capitalism lens. A firm contemplating heavy investment in transferable worker skills faces a coordination problem: if it trains workers in skills usable elsewhere, competitors who don't invest can poach the trained staff, and workers won't participate unless assured the new skills will be appropriately remunerated. The passages show the firm will only commit if surrounding institutions — operationalized through the state, employer associations, and unions — provide four coordinating functions: information circulation, deliberation, monitoring, and sanctioning. In a 'Rhineland' economy like Germany or the Netherlands, these institutional complementarities exist, so firms sustain occupational labor markets with skill-based job titles and institutionalized training (per Marsden 1986). In an Anglo-American setting lacking that coordination, the same firm rationally opts for buy-over-make, running with higher vacancies and heavier workloads for existing staff. The work structure differs not because the skill is more or less unique, but because the institutional supply of ready-trained labor differs.

• If a job category's skill is generic and available ready-trained in the external labor market, then lean toward buy/contract — but if operational efficiency and reputational effects depend on those incumbents' daily motivation and cooperation, then retain and manage them beyond mere rule-compliance.
• If you are designing service work, then explicitly reconcile the operations-efficiency logic with the marketing-customer-satisfaction logic before finalizing job design, because unresolved tension pushes conflicting demands onto workers.
• If considering large investments in transferable skills, then first confirm the institutional coordination (information, deliberation, monitoring, sanctioning) exists; if it doesn't, expect under-investment and design around bought-in or firm-specific skills instead.
• If your workers are knowledge workers whose work is central to the firm, then expect distinctive and multiple identities and aspirations that may not match the employer's, and design work and voice systems accordingly rather than assuming alignment.
• If operating in the public sector, then treat institutional patterns as embedded and hard to change, and temper 'new public management' efficiency reforms against durable institutional behavior.

• Taking the individual firm as the unit of analysis. The tell: your work-design conclusions don't change across manufacturing, services, or public sector, and you've ignored sectoral and occupational context that materially alters what's possible.
• Treating low-strategic-value labor as a pure commodity per an HR architecture reading. The tell: you specify only rules and regulations for those roles and are then surprised by motivation, cooperation, or reputational failures in the production chain.
• Optimizing job categories in isolation. The tell: each role looks efficient on paper but hand-offs and coordination between categories break down, because you focused on the value of skill and not the interactions between jobs.
• Assuming firms will invest in transferable training just because it's beneficial. The tell: your training scheme collapses through poaching or non-participation, because the coordinating institutions (state, associations, unions) weren't there to provide credible assurance.
• Letting one function's logic silently dominate service design. The tell: operations quietly wins on cost while customer satisfaction erodes (or vice versa), because the operations/marketing tension was never surfaced and negotiated in the work structure.

• ▸Sector/occupational context: name the sector (manufacturing/service/knowledge/public) and its dominant constraints
• ▸Work system archetype: classify by work content characteristics and note the matching job design strategy
• ▸Job category & make-buy call: for each category, state insource vs. contract and the reason
• ▸External labor availability: is ready-trained labor available in the market? (yes/no + evidence)
• ▸Cross-category interactions: list the coordination dependencies between this role and others in the production chain
• ▸Motivation/cooperation mechanisms: daily levers to keep incumbents working in the client's interest beyond rule-compliance
• ▸Institutional coordination check: are information, deliberation, monitoring, sanctioning present for any transferable-skill investment? (yes/no)
• ▸Efficiency vs. satisfaction tension: name the operations/marketing conflict this design creates and how it's resolved
• ▸Economic vs. legitimacy balance: state how the design balances competitiveness/efficiency against justice/legitimacy objectives

1. 1Start from strategy: treat intended HR policies as 'the pattern of planned human resource deployments and activities intended to enable the firm to achieve its goals' (Wright and McMahan 1992). Map current firm goals before selecting practices.
2. 2Make the make-or-buy decision first. As Kaufman (2004) argues, the selection of HR practices presupposes a prior choice between market and hierarchy; decide which capabilities you build internally versus buy externally, weighing the transaction costs of managing, negotiating, and respecifying contracts against internal costs.
3. 3Design across all three AMO levers together: Ability (selection, training), Motivation (rewards, performance management), and Opportunity to contribute (employee involvement, job design). Draw on the work-content taxonomy and job design strategies rather than picking practices in isolation.
4. 4Build in synergy — check internal fit within the HRM architecture so practices reinforce rather than contradict one another (e.g. rewards that support, not undercut, the involvement and job design you have chosen).
5. 5Specify the causal chain. Lay out mediators — job satisfaction, trust-in-management, attitudinal commitment, discretionary behavior — so you can state how each policy is intended to affect performance, not just assume it.
6. 6Assign enactment to line managers explicitly. Intended policies only work when 'effectively enacted by line managers'; name who delivers each practice and how.
7. 7Justify policies on two grounds, not one: the business case (commercial outcomes) and social legitimacy. Avoid defending every policy solely by the profit-oriented bottom line.
8. 8Review formally and periodically. Test relative performance against alternative investments (technology, advertising, property) rather than assuming a new technique inevitably improves the business.

Consider a firm deciding whether to protect learning and cooperative activity in-house. Following Kaufman (2004), management first faces the make-or-buy choice: outsourcing certain work appears cheaper on cost-per-unit-output. But qualitative HRM evidence shows outsourcing carries costs to worker morale and commitment (George 2003; Logan et al. 2004) that do not fit neat cost models. Teece's (2002) dynamic-capabilities reasoning warns that using market benchmarks may reduce the value attached to firm-specific activities — 'many distinct elements of internal organization simply cannot be replicated in the market' (2002:158). So senior management chooses hierarchy for these activities and designs an intended AMO bundle: selective hiring and training (Ability), performance management plus rewards tuned to retention (Motivation), and employee involvement with job design that supports cooperation (Opportunity). They then specify the intended chain — how these raise commitment and discretionary behavior — and assign delivery to line managers.

• If the activity is firm-specific, learning-intensive, or cooperative, then keep it inside a hierarchy and design AMO policies around it; market benchmarks will erode its value (Teece 2002).
• If a policy can only be justified by the bottom line and adds nothing to social legitimacy, then treat that justification as incomplete — HRM also serves organizational needs for legitimacy (Lees 1997; Gooderham et al. 1999).
• If a practice is not effectively enacted by line managers or does not enhance employee attitudes, then it will not produce performance regardless of its design (Purcell et al. 2003).
• If two chosen practices pull in opposite directions (e.g. individual incentives against team involvement), then rework the bundle until internal fit holds.
• If you cannot articulate the cause–effect chain and mediators for a policy, then you cannot yet claim it contributes to competitive advantage — spell out how HRM creates valuable capabilities and barriers to imitation (Boxall and Purcell 2003; Wright et al. 2003).

• Means-become-ends displacement: staff treat the target as the goal ('We have fulfilled our quota of job upgradings for this month, why should we do any more?'), undermining the purpose the policy was designed for. Tell: activity metrics hit while intended outcomes stagnate.
• Bottom-line tunnel vision: justifying every HR policy purely in profit terms, misunderstanding the plurality of organizational effectiveness. Tell: no attention to social legitimacy or employee experience.
• Best-practice naivety: assuming managers are 'waiting with bated breath' to implement the latest technique, ignoring the economics of production in their sector. Tell: a new practice adopted by mimicry with no fit to context.
• Intended-but-not-enacted gap: policies designed at the top that line managers never deliver, so employee attitudes are untouched. Tell: policy exists on paper but employees report no change.
• Weak measurement and no mediators: measuring presence of practices crudely and never modeling how they affect satisfaction, trust, and discretionary behavior (Gerhart et al. 2000). Tell: performance claims with no cause–effect chain to support them.

• ▸Firm goal targeted — state the specific organizational goal this policy set enables
• ▸Make-or-buy decision — record which capabilities are built internally vs bought, with cost/commitment reasoning
• ▸Ability practices — list selection and training practices chosen
• ▸Motivation practices — list rewards and performance management practices chosen
• ▸Opportunity practices — list employee involvement and job design choices
• ▸Internal fit check — note how the practices reinforce rather than contradict each other
• ▸Intended causal chain — name the mediators (satisfaction, trust, commitment, discretionary behavior) linking policy to performance
• ▸Line-manager enactment — assign who delivers each practice and how it is supported
• ▸Dual justification — record the business case and the social-legitimacy rationale

1. 1Separate the three states of every practice before you measure anything: intended (what senior management designed and put in the HR manual/web pages), actual (what line managers really apply), and perceived (what employees experience and judge through a lens of fairness). The gap between espousal and enactment lives at the line-manager layer.
2. 2Map each intended practice to the AMO dimensions — ability, motivation, opportunity to participate — then audit how the line manager actually delivers each one. Enactment failure usually shows up in the motivation and opportunity dimensions before ability.
3. 3Assess the line manager as a mediator, not a conduit. Following Truss (Hewlett Packard), treat the manager as a 'powerful mediator between the individual and HR practices' — measure the quality of the dyadic relationship (leader–member exchange), not just whether the policy was communicated.
4. 4Measure the climate the manager produces: levels of trust (Whitener), work climate, and job experience (pace, effort, autonomy, challenge, stress). These perceived-practice signals are what actually feed attitudinal outcomes.
5. 5Control for the wider variables that co-vary with enactment — leadership, organizational culture, operational strategies (staffing levels, job design, social relations of production) — because HR practice measures can act as proxies for these. Wright et al. (2005) warn that without a full set of controls you cannot make valid causal inferences.
6. 6Trace the causal chain forward: enactment/leadership → perceived practices → attitudinal outcomes (job satisfaction, commitment, willingness to cooperate) → behavioral outcomes (discretionary/OCB, attendance, retention) → lagged organizational effectiveness. Diagnose breakdowns by walking backward from the behavioral outcome to the enactment step.

Truss (2001) studied Hewlett Packard using both qualitative and quantitative methods and found that managers act as a powerful mediator between the individual and HR practices. Consider an intended appraisal policy designed by senior management and documented in the HR manual to build motivation and opportunity to participate. Two line managers apply the same policy: one runs it as a genuine developmental conversation that raises trust and job satisfaction; the other treats it as a box-ticking ritual, so employees perceive the same intended practice through a lens of unfairness. The perceived practice diverges sharply even though the intended practice is identical — the difference is enactment. In HP this mediation showed up in both interview data and statistical measures, confirming the manager as the hinge between policy and experienced practice. The lesson: the espousal–enactment gap (Hutchinson and Purcell 2004) is a line-management phenomenon, and measuring only intended practices misattributes performance effects.

• If intended and perceived practices diverge for the same policy, then the cause is enactment — investigate the line manager's delivery and the LMX quality, not the policy design.
• If a 'high-performance' practice fails to lift performance, then check inter-personal relations first — Delery and Shaw note that success of such practices hinges on the nature of interpersonal relations in the organization.
• If you cannot measure leadership, culture, and manager behavior, then you cannot control for them — and per Wright et al. any HR-to-performance correlation you report risks being spurious.
• If employees show low trust or willingness to cooperate despite well-designed policies, then treat it as a perceived-practice and climate problem transmitted by the manager, not a policy gap.
• If a behavioral outcome (absence, turnover, low discretionary behavior) is weak, then trace back through attitudinal outcomes to the perceived practice and the enacting manager rather than revising the intended policy.

• Measuring only intended practices from the HR manual and assuming they equal what employees experience — the tell is a strong policy audit paired with poor employee attitude data.
• Treating the line manager as a passive transmitter of policy — the tell is that identical policies produce different outcomes across units and you have no manager-level variable to explain it.
• Omitting line managers as HRM agents entirely, which Delery and Shaw call 'curious' — the tell is a causal model that jumps from HR practices straight to performance with no mediation layer.
• Reporting an HR–performance correlation without controlling for leadership, culture, and operational strategy — the tell is a 'spurious' relationship where HR measures are silently proxying for manager behavior.
• Ignoring perceived climate signals like trust and job experience — the tell is high scores on formal practice provision but low morale, cooperation, and discretionary behavior.

• ▸Policy name: the specific intended HR practice as written in the manual/web pages
• ▸Intended AMO target: which of ability, motivation, opportunity this policy is designed to affect
• ▸Actual enactment: what the line manager really does when applying it
• ▸Perceived practice: how employees describe and judge it (fairness/justice lens)
• ▸Enactment gap: one-line note on the divergence between intended, actual, and perceived
• ▸LMX/trust indicator: current quality of manager–employee relationship and trust level
• ▸Climate signal: observed job experience — pace, effort, autonomy, challenge, stress
• ▸Attitudinal outcome affected: job satisfaction, commitment, or willingness to cooperate
• ▸Corrective owner + action: who fixes the enactment (manager coaching vs. policy redesign) and next step

1. 1Separate what you intend from what employees experience: audit your written HR policies against how line managers actually enact them day to day, because perceived practices are shaped by managers and climate, not by the policy document.
2. 2Trace the causal chain deliberately — intended HR practice, then implemented practice, then perceived practice, then employee attitude/behavior, then outcome (following the 'Revised HR causal chain' and 'people management, HRM, and organizational effectiveness' logic). Do not skip the perception link when explaining results.
3. 3Sample perceptions at the individual employee level, not the HR-department level: ask employees what they experienced (was I trained, was I consulted, was I paid fairly), since two people under the same formal policy can report different experiences.
4. 4Assess line-manager mediation directly — measure whether managers deliver the practice consistently, because inconsistent enactment is where intended and perceived practices diverge.
5. 5Distinguish rhetoric from reality explicitly: name where the espoused HRM message ('high commitment') may be experienced as its opposite (intensification), and check employee reports for that gap.
6. 6Read perceived practices against the strategic-uncertainty logic: where employee groups are perceived as low strategic uncertainty, expect direct-control practices to be experienced; where high, expect high-commitment practices — and check whether employees perceive the version you claim to run.
7. 7Aggregate perception data into the black box between HRM and performance: use it to explain why management initiatives enlist employee support or fail to do so.

Purcell and Kinnie's 'black box' problem shows the construct at work: a firm's formal HR policy lists high-commitment practices — training, voice, fair pay — yet performance gains fail to appear. The intended practices (chapter 26's causal chain) are real on paper, but employees' perceptions diverge because line managers enact them unevenly. One team's manager runs regular voice sessions and coaches; another's treats the same policy as paperwork. When you survey employees rather than the HR function, the first team reports being consulted and developed while the second reports the Braverman experience — 'manipulation, pacification, and adjustment.' The measured practice that predicts attitude and behavior is the perceived one, which is why Purcell's revised chain inserts people management (manager enactment) between HRM and organizational effectiveness. The lesson: the policy did not fail; its perception, mediated by managers and climate, split.

• If two employees under one policy report different experiences, then treat the line manager and local climate as the unit of analysis, not the policy.
• If measured performance effects are absent despite strong formal practices, then measure perceived practices next — the break is likely inside the black box.
• If employee constituencies are perceived as low strategic uncertainty, then expect direct-control practices to dominate experience regardless of high-commitment rhetoric.
• If espoused HRM language emphasizes commitment while employees report intensification, then read the gap as the rhetoric–reality distinction and investigate enactment, not messaging.
• If you can only survey the HR department, then you are not measuring perceived practices — defer conclusions until you reach employees.

• Measuring intended practices and calling them experienced ones. Tell: your data comes from HR records or policy documents, never from employees themselves.
• Ignoring line-manager mediation and attributing outcomes straight to policy. Tell: identical policies produce inconsistent results and you have no manager-level variable to explain it.
• Taking HRM rhetoric at face value as employee reality. Tell: your report echoes the espoused 'high commitment' claim while employee comments describe control and intensification.
• Averaging perceptions to the firm level and losing the variation that matters. Tell: a single satisfaction score masks teams with opposite experiences.
• Treating perception as 'too soft' to measure rigorously (the Lazear objection). Tell: you drop the perception link entirely because it resists unambiguous answers, leaving the black box unopened.

• ▸Practice named: the specific HR practice under review (e.g. training, voice, pay).
• ▸Intended version: what the written policy promises, one line.
• ▸Enacted version: what the line manager actually delivers, from observation.
• ▸Perceived version: how employees report experiencing it, in their words.
• ▸Line-manager mediation note: where and how the manager alters delivery.
• ▸Climate signal: the local cue (trust, control) coloring the perception.
• ▸Gap flag: rhetoric–reality mismatch — yes/no and where.
• ▸Uncertainty read: is this group perceived as low or high strategic uncertainty?
• ▸Action: the single change to close the intended–perceived gap.