Theory Development & Hypotheses (smj-theory-development)
When to trigger
- Your hypotheses read as signed predictions ("X is positively related to Y") with no logical engine
- The theory section restates prior findings instead of building a new argument
- A moderator is added but you cannot explain why it changes the effect
- The mechanism connecting strategic choice to performance is implicit or hand-waved
What SMJ means by "theory"
SMJ wants a causal logic: a chain of premises that explains why a strategic choice or condition produces a performance/advantage consequence, anchored in a recognized strategy perspective. A hypothesis without a stated mechanism is a correlation in disguise. The reviewers expect you to show the logical steps, not just the conclusion.
Choosing and combining lenses
| Lens |
Core engine |
Typical use |
| Resource-based / capabilities |
Heterogeneous, hard-to-imitate resources earn rents (Wernerfelt, 1984, SMJ) |
Why advantage persists |
| Dynamic capabilities |
Sensing/seizing/reconfiguring under change (Teece, Pisano & Shuen, 1997, SMJ) |
Adaptation, renewal |
| Transaction cost economics |
Governance minimizes contracting hazards |
Make-vs-buy, alliances, scope |
| Agency / governance |
Incentive misalignment shapes strategic choice |
Boards, ownership, M&A |
| Competitive dynamics |
Action–response, awareness–motivation–capability |
Rivalry, market entry |
| Behavioral theory of the firm |
Aspirations, search, attention |
Risk taking, R&D search |
| Real options |
Sequential investment under uncertainty |
Entry, staging, divestiture |
| Institutional / nonmarket |
Legitimacy and regulatory environment shape advantage |
CSR, political strategy |
Borrowing a lens is fine; extending or qualifying it is what earns a contribution. State which lens you build on and exactly where you depart from it.
Anchor RBV with Wernerfelt (1984), SMJ when you want an SMJ-internal reference. Barney's much-cited 1991 RBV statement appeared in Journal of Management, not SMJ — do not list it as an SMJ landmark.
SMJ is a theory-develop-and/or-test journal: unlike AMR, it does not publish purely conceptual papers, so your theory must set up implications that are evaluable (testable, or otherwise assessable for qualitative/formal work).
Building the argument (mechanism-first)
- Establish the baseline relationship. State the focal X → Y and the strategy logic for its sign.
- Specify the mechanism. Name the mediating process (e.g., reduced imitation, lower coordination cost, shifted aspirations). The mechanism is the contribution's spine.
- Develop moderators as theory, not robustness. A moderator hypothesis must explain how the boundary condition strengthens or reverses the mechanism — not merely "the effect is bigger for big firms."
- Surface and address the tension. SMJ rewards arguments that resolve a real tension or counter-prediction (e.g., a resource that both enables and constrains).
- State the counterfactual and scope conditions. Where does the logic hold, and where does it break?
Hypothesis craft
- Each H states direction and carries a one-line mechanism in the surrounding text.
- Order hypotheses so the baseline (H1) precedes mediation/moderation (H2, H3…).
- Avoid "kitchen-sink" hypothesizing; 2–4 tightly linked hypotheses beat seven loose ones.
- If you theorize a mediator, plan a mechanism test in
smj-data-analysis (mediation alone is weak evidence).
Checklist
Anti-patterns
- "X is positively associated with Y" with no mechanism — the most common SMJ theory weakness
- Hypothesis salad: many predictions, no integrating logic
- Re-deriving an established result and calling it theory
- A moderator with no theoretical reason for changing the effect
- Mechanism claimed in theory but never tested empirically
- Using a strategy lens as decoration while the real argument is generic
Output format
【Lens】RBV | dynamic capabilities | TCE | agency | competitive dynamics | BTOF | real options | institutional
【Gap/tension resolved】...
【Mechanism (one sentence)】X affects Y because ...
【Hypotheses】H1 (baseline) / H2 (mediation) / H3 (moderation) — each with its mechanism
【Boundary conditions】...
【Empirical implication】mechanism test needed → flag for smj-data-analysis
【Next step】smj-literature-positioning
Templates & resources
Source: brycewang-stanford/Awesome-Journal-Skills → Strategic-Management-Journal-Skills/skills/smj-theory-development/SKILL.md
1---2name: smj-theory-development3description: Use when building the theoretical mechanism and hypotheses for a Strategic Management Journal (SMJ) manuscript. Constructs the argument and develops hypotheses; it does not position the literature or run estimation.4---567# Theory Development & Hypotheses (smj-theory-development)89## When to trigger1011- Your hypotheses read as signed predictions ("X is positively related to Y") with no logical engine12- The theory section restates prior findings instead of building a new argument13- A moderator is added but you cannot explain *why* it changes the effect14- The mechanism connecting strategic choice to performance is implicit or hand-waved1516## What SMJ means by "theory"1718SMJ wants a **causal logic**: a chain of premises that explains *why* a strategic choice or condition produces a performance/advantage consequence, anchored in a recognized strategy perspective. A hypothesis without a stated mechanism is a correlation in disguise. The reviewers expect you to show the logical steps, not just the conclusion.1920## Choosing and combining lenses2122| Lens | Core engine | Typical use |23|-------------------------------|--------------------------------------------------------|-------------|24| Resource-based / capabilities | Heterogeneous, hard-to-imitate resources earn rents (Wernerfelt, 1984, SMJ) | Why advantage persists |25| Dynamic capabilities | Sensing/seizing/reconfiguring under change (Teece, Pisano & Shuen, 1997, SMJ) | Adaptation, renewal |26| Transaction cost economics | Governance minimizes contracting hazards | Make-vs-buy, alliances, scope |27| Agency / governance | Incentive misalignment shapes strategic choice | Boards, ownership, M&A |28| Competitive dynamics | Action–response, awareness–motivation–capability | Rivalry, market entry |29| Behavioral theory of the firm | Aspirations, search, attention | Risk taking, R&D search |30| Real options | Sequential investment under uncertainty | Entry, staging, divestiture |31| Institutional / nonmarket | Legitimacy and regulatory environment shape advantage | CSR, political strategy |3233Borrowing a lens is fine; **extending or qualifying** it is what earns a contribution. State which lens you build on and exactly where you depart from it.3435> Anchor RBV with **Wernerfelt (1984), SMJ** when you want an SMJ-internal reference. Barney's much-cited 1991 RBV statement appeared in *Journal of Management*, **not** SMJ — do not list it as an SMJ landmark.3637SMJ is a **theory-develop-and/or-test** journal: unlike AMR, it does not publish purely conceptual papers, so your theory must set up implications that are evaluable (testable, or otherwise assessable for qualitative/formal work).3839## Building the argument (mechanism-first)40411. **Establish the baseline relationship.** State the focal X → Y and the strategy logic for its sign.422. **Specify the mechanism.** Name the mediating process (e.g., reduced imitation, lower coordination cost, shifted aspirations). The mechanism is the contribution's spine.433. **Develop moderators as theory, not robustness.** A moderator hypothesis must explain how the boundary condition *strengthens or reverses the mechanism* — not merely "the effect is bigger for big firms."444. **Surface and address the tension.** SMJ rewards arguments that resolve a real tension or counter-prediction (e.g., a resource that both enables and constrains).455. **State the counterfactual and scope conditions.** Where does the logic hold, and where does it break?4647## Hypothesis craft4849- Each H states direction *and* carries a one-line mechanism in the surrounding text.50- Order hypotheses so the baseline (H1) precedes mediation/moderation (H2, H3…).51- Avoid "kitchen-sink" hypothesizing; 2–4 tightly linked hypotheses beat seven loose ones.52- If you theorize a mediator, plan a mechanism test in `smj-data-analysis` (mediation alone is weak evidence).5354## Checklist5556- [ ] Every hypothesis is preceded by an explicit causal mechanism57- [ ] You name the lens you extend and the specific gap/tension you resolve58- [ ] Moderators are theorized (why the mechanism shifts), not bolted on59- [ ] A counter-argument / competing prediction is acknowledged and addressed60- [ ] Scope and boundary conditions are stated61- [ ] The argument predicts something a reader could not already infer from prior work6263## Anti-patterns6465- "X is positively associated with Y" with no mechanism — the most common SMJ theory weakness66- Hypothesis salad: many predictions, no integrating logic67- Re-deriving an established result and calling it theory68- A moderator with no theoretical reason for changing the effect69- Mechanism claimed in theory but never tested empirically70- Using a strategy lens as decoration while the real argument is generic7172## Output format7374```75【Lens】RBV | dynamic capabilities | TCE | agency | competitive dynamics | BTOF | real options | institutional76【Gap/tension resolved】...77【Mechanism (one sentence)】X affects Y because ...78【Hypotheses】H1 (baseline) / H2 (mediation) / H3 (moderation) — each with its mechanism79【Boundary conditions】...80【Empirical implication】mechanism test needed → flag for smj-data-analysis81【Next step】smj-literature-positioning82```8384## Templates & resources8586- [`../../resources/official-source-map.md`](../../resources/official-source-map.md) — SMJ scope and verified landmark strategy-theory papers (Wernerfelt 1984; Teece, Pisano & Shuen 1997)8788---8990**Source:** [`brycewang-stanford/Awesome-Journal-Skills`](https://github.com/brycewang-stanford/Awesome-Journal-Skills) → `Strategic-Management-Journal-Skills/skills/smj-theory-development/SKILL.md`