# Methodology

## 1. Scope

The Prayer Index and Fasting Index trace the relative compliance with each spiritual discipline in Western Christianity from approximately 100 CE through 2025 CE. Both are normalized to a 0–100 scale where 100 represents peak historical compliance and 0 represents complete absence.

## 2. Index construction

### 2.1 Modern era (1939–present): direct survey data

For the modern era, survey data is available and the indices are anchored to it:

- **Gallup** church attendance and prayer-frequency data, beginning 1939.
- **Pew Research Center** Religious Landscape Study (2007, 2014, 2023–24) provides daily prayer frequencies.
- **Barna Group** tracking of weekly prayer (since 1996), Bible engagement, religious identification.
- **Lifeway Research** Lent observance surveys (2016, 2024).
- **NBER Working Paper 34060** (Barro et al., 2025) cross-national long-term religious attendance via ISSP retrospective data.

Where survey instruments differ in question wording or denominator, the index uses scholarly cross-tabulations and the most consistent measure available. Confidence is marked HIGH for this era.

### 2.2 Pre-modern era (100–1938): inference from canonical and historical sources

Direct survey measurement of religious practice did not exist before 1939. For the pre-modern era, the indices are inferred from:

- **Canonical requirements** — what the Catholic Church (and later Protestant churches) officially required for fasting and prayer practice. Major sources:
  - Didache (~90 CE)
  - Rule of St. Benedict (529)
  - Fourth Lateran Council (1215)
  - 1917 Code of Canon Law
  - 1983 Code of Canon Law
- **Church-attendance estimates** from scholarly historical research (Finke & Stark, Putnam, Bynum).
- **Patristic and monastic sources** documenting actual practice in specific eras.
- **Black Death / Reformation / Enlightenment scholarship** providing broad turning-point estimates.

Confidence is marked MEDIUM or LOW for pre-modern values, with appropriate error bars.

### 2.3 Peak calibration

The peak (100 on the scale) is set at approximately 1300 CE, immediately before the Black Death. Scholarly consensus places this period as the institutional and demographic peak of Western Christendom: Fourth Lateran Council requirements were fully institutionalized, mendicant orders had spread popular devotion, and church attendance was near-universal by social and legal expectation.

Setting the peak in 1300 means the modern era (2023: prayer 48, fasting 20) is compared not to its own historical baseline but to medieval peak. This is a defensible analytical choice but it can make the modern values look more dramatic than they would on a peak-of-survey-era baseline.

## 3. What "prayer" means in this dataset

Prayer is measured as approximately "regular, intentional addressing of God by Christians in the relevant population." This intentionally bundles:

- Liturgical / public prayer (Mass, Divine Office, hymns)
- Private daily prayer
- Family prayer at meals
- Personal devotional practice

The index does not distinguish between these forms; the medieval Catholic peasant attending Mass, the Reformation Lutheran reading the catechism with their family, and the modern American praying privately before sleep all count as "practicing prayer."

## 4. What "fasting" means in this dataset

Fasting is measured as approximately "compliance with the dominant fasting discipline of the era for the relevant population."

The dominant discipline has changed dramatically:

- **Pre-1517 (universal Catholic discipline)**: 40 days of Lent + Advent fast + Ember Days + Vigils + Wednesday/Friday abstinence year-round. Approximately 120+ fasting days per year for an observant Catholic, with strict food rules (no meat, dairy, eggs on certain days; one meal per day in late afternoon/evening; specific timing).

- **1517–1965 (split discipline)**: Protestants make fasting voluntary; Catholics maintain strict discipline through 1917 Code, with progressive relaxations in 1741 (meat allowance), 1886 (Sunday Lent meat/eggs/dairy), 1917 (Advent fast removed).

- **1966–present (minimal Catholic discipline)**: *Paenitemini* (Pope Paul VI, 17 February 1966) reduces mandatory Catholic fasting to two days/year: Ash Wednesday and Good Friday. Friday abstinence outside Lent made optional in 1966 (US bishops). Most Protestant traditions have no required fasting at all. Modern practice is largely "giving something up for Lent" (varied: chocolate, social media, alcohol) rather than strict food abstinence.

The index measures compliance with whatever the dominant discipline of the era was. A modern Catholic who fasts only on Ash Wednesday and Good Friday scores 100% on the modern discipline but represents perhaps 5% of historical Catholic fasting practice.

## 5. The 1966 inflection — methodologically

The Fasting Index drops from 90 (1955) to 50 (1966) — a 40-point single-year drop — to reflect the institutional discipline change. This is the largest single-year shift in the entire dataset. It is supported by:

- The text of *Paenitemini* itself (a verbatim Apostolic Constitution change).
- Subsequent measurement of Catholic Mass attendance collapse (75% weekly in 1955 → 35% among young adults by 1975 per Gallup historical analysis).
- Lifeway 2024 data showing only 26% of Americans observe Lent, with only 50% of those reporting food-related fasting (i.e., ~13% of Americans practice any meaningful food-fasting).

Whether this drop represents a "decline" in fasting (the institutional rules were lowered; individual compliance with new rules is high) or a "collapse" (the underlying practice that defined Catholic identity for 1900 years largely ended) is partly a definitional question. The report uses the historical-discipline framing — i.e., the practice of strict fasting as it was understood for 19 centuries largely ended.

## 6. Confidence stratification

| Era | Confidence | Justification |
|---|---|---|
| 1939–present | HIGH | Direct survey data from Gallup, Pew, Barna, Lifeway |
| ~1500–1938 | MEDIUM | Documented canonical requirements + scholarly church-attendance estimates |
| 100–1500 | LOW | Inference from patristic sources, monastic rules, historical accounts |

Pre-1500 values should be read with ±10 point error bars; pre-500 values with ±15 point error bars. The modern era values are within ±3 points.

## 7. What this methodology produces

- A descriptive trajectory of two foundational spiritual disciplines across two millennia.
- A documented institutional cause for the 1966 fasting collapse.
- A comparison of two practices that helps explain why some forms of religious decline are gradual and others are sudden.
- An honest accounting of which periods have direct measurement and which require inference.

## 8. What this methodology does not produce

- A causal model of why Christianity is declining.
- A prediction of where the trajectories will go next.
- A normative judgment of the 1966 reforms (they had a stated theological rationale — that penance should be voluntary and interior rather than externally compelled).
- A comparable trajectory for non-Christian fasting/prayer traditions (which have their own distinct histories).