eursr-theory-building

Theory & Argument Building (eursr-theory-building)

At ESR a finding becomes a contribution when it is attached to an **explicit mechanism that other

sociologists can use and stated as testable hypotheses** the design can adjudicate. ESR's house

form is often a macro–micro argument: how an institutional or contextual feature shapes an

individual-level social process. This skill turns data into theory.

When to trigger

• The models are strong but the "so what / why" is thin
• A reviewer said the paper is "atheoretical," "variable-driven," or "hypotheses tacked on"
• You need to state the mechanism, the hypotheses, and the scope conditions
• Reconciling a cross-national pattern with an individual-level explanation

Build the argument

1. Concept. Define the key construct (the form of inequality, mobility, integration, the attitude)

and distinguish it from neighbors; say how it is measured in principle.

1. Mechanism. The social process generating the pattern — who acts, why, under what structural or

institutional conditions. Make the macro–micro link explicit (how context C changes the

individual calculus or constraint).

1. Hypotheses. Derive directional, testable predictions before the analysis — including

cross-level interactions (does the country/context feature moderate the individual effect?).

1. Scope conditions. The contexts, populations, periods, and institutional regimes where the

argument holds — essential for a comparative claim.

The macro–micro template (ESR-specific)

State the argument as a chain a reviewer can test:

Context feature (macro)  →  changes individual constraint/incentive (micro)  →  observable outcome
e.g., earlier tracking   →  earlier high-stakes choice under uncertainty      →  larger secondary
                            for lower-SES families                              (choice-based) SES gap

The cross-level interaction (macro × micro) is usually the hypothesis that carries an ESR paper —

it is what cross-national or panel data can identify and a single cross-section cannot.

The "portability" test (ESR-specific)

Ask: Could a sociologist studying a different institution or country import this mechanism? If yes,

you have a contribution. If it only works for your wave or country, generalize the logic or reframe

(back to eursr-topic-selection).

What "theoretical contribution" means to an ESR referee

ESR referees grade portability and the macro–micro logic, not local cleverness or model fit alone.

| Referee judgment | A passing argument shows | A failing argument shows |

|------------------|---------------------------|---------------------------|

| "Atheoretical / variable-driven" | a named mechanism + derived hypotheses | a list of significant predictors |

| "Hypotheses tacked on" | predictions stated before the analysis | hypotheses that restate the results |

| "Why does context matter?" | an explicit macro–micro / cross-level link | a country dummy with no mechanism |

| "Where does it stop?" | scope conditions across regimes | an unbounded claim |

Worked micro-example (illustrative)

A panel study of unemployment scarring is built from data to a portable claim.

Concept: scarring = a lasting wage/employment penalty from a past unemployment spell, net of selection
Mechanism (macro–micro): generous, activating welfare regimes shorten spells and signal less, so the
  individual scar is smaller where re-employment support is strong
Hypothesis (cross-level): the scarring penalty (micro) is weaker in high-activation regimes (macro)
  → a macro × micro interaction the comparative panel can test
Scope conditions: prime-age workers; regimes with active labor-market policy; pre-/post-crisis
Portability: a scholar of any state-buffered risk (illness, divorce) can import the buffering logic

The data yield a mechanism and a cross-level prediction another comparative subfield can use — the move ESR rewards.

Referee pushback → ESR-specific fix

• "This is a regression with a story attached." → Derive the hypotheses from the mechanism before the

models; show the predictions are not just the estimates restated.

• "The country variable does no theoretical work." → Replace the dummy with a measured institutional

feature and a cross-level interaction the data can identify.

• "Where does the argument hold?" → Add explicit scope conditions across regimes and periods.

Calibration anchors

• The cross-level interaction is usually the contribution. ESR's comparative/longitudinal leverage

exists to test how context moderates an individual process — make that the headline hypothesis.

• Theory before tests. State the mechanism and derive hypotheses first; theorizing after the results

reads as HARKing to expert referees.

• State the claim plainly with bounded scope. A mechanism a comparative sociologist can restate in a

sentence, with explicit scope conditions, outperforms one buried under model output.

Anti-patterns

• A finding with no mechanism ("X predicts Y" and stop)
• Hypotheses that merely restate the coefficients (HARKing)
• A country/context dummy standing in for a theorized institutional feature
• Grand claims with no scope conditions
• Burying the argument under tables — the contribution must be stated plainly

Output format

【Core claim】one sentence
【Mechanism】the macro–micro / causal logic
【Key concept(s)】defined and distinguished
【Hypotheses】directional predictions incl. the cross-level interaction → research-design
【Scope conditions】regimes / periods / populations where it holds
【Portability】who else can use this argument
【Next】eursr-research-design

Supplementary resources

• [../../resources/external_tools.md](../../resources/external_tools.md) — multilevel / SEM / panel tooling across methods
• [../../resources/official-source-map.md](../../resources/official-source-map.md) — ESR scope and contribution expectations