Open data: UK macroeconomic series from 1086 to present. Bank of England Millennium of Macroeconomic Data. (Year)

Rows: 22,487
Columns: 6

Hundreds of annual time series covering nominal and real GDP, CPI, wages, interest rates, public debt, money supply, and equity markets for the UK. Some series go back to 1086. Most are continuous from at least 1700.

AI analysis

Dataset Analysis: Bank of England Millennium of Macroeconomic Data

1. Dataset Overview & Research Value

This dataset represents one of the most historically comprehensive macroeconomic compilations publicly available, spanning nearly 930 years of UK economic history (1086–2016) across hundreds of annual time series. Published by the Bank of England, it covers the full breadth of macroeconomic measurement: nominal and real GDP, consumer prices, wages, interest rates, public debt, money supply, and equity markets. The extraordinary temporal depth makes it uniquely valuable for studying long-run structural relationships — such as the historical link between money supply growth and inflation, or the co-movement of interest rates and public debt across war cycles, industrial revolutions, and financial crises. No source URL was provided, but the Bank of England is the authoritative publisher, and coverage through 2016 suggests the dataset may require supplementation for more recent analysis.

2. Data Quality Observations

Data quality is notably strong for a dataset of this age and complexity. Total null cells across all 6 columns register at zero, which is impressive but warrants some interpretive caution — in long historical series, zeros or placeholder values sometimes substitute for true missing data, and this should be verified on inspection. No type mismatches are flagged. Duplicate row counts are pending Phase D recomputation and should be confirmed before any aggregation. The 65 distinct variableid values versus 63 distinct variable names suggests a minor labeling inconsistency worth investigating, as two variable IDs may share display names or vice versa — a small but meaningful quality flag for programmatic joins.

3. Key Column Distributions

The value column tells the most analytically interesting story. The mean (38,248.6) sits far above the median (40.03), a dramatic divergence that reflects the mixture of vastly different unit scales across series — index values near 1.0 coexist with nominal GDP figures in the hundreds of thousands. The interquartile range (Q1 = 1.82, Q3 = 2,518.65) confirms that the bulk of observations are relatively small-valued, while the skewness of 6.7 and 4,372 flagged outliers (~19.4% of rows) are almost certainly artifacts of mixing units rather than true data anomalies. The unit column (31 distinct values) is therefore critical context for any analysis — values should never be compared or correlated across different units without normalization. The year column shows 931 distinct years against a ~930-year span, indicating near-complete annual coverage with minimal gaps.

4. Recommended Join Key Columns

The year column is the natural and primary join key for cross-dataset correlation, as it is the sole temporal anchor and carries zero nulls with consistent annual granularity. For within-dataset pivoting and subsetting, variableid is preferable over variable as a programmatic key given the slight name-count discrepancy noted above. section (7 distinct values) serves as a useful segmentation key to isolate thematic subsets (e.g., monetary, fiscal, real economy) before joining, reducing noise from mixing unrelated series. When joining externally, standardizing year to a plain integer or ISO year format will improve compatibility with most partner datasets.

5. Recommended Pairing Datasets for Correlation Discovery

Several dataset categories would pair powerfully with this data. European or global historical macroeconomic series (e.g., Maddison Project GDP data, or historical series for France, Netherlands, and the US) would enable comparative analysis of industrialization timing and financial crisis contagion. UK political and institutional history datasets — covering parliament composition, war years, or colonial trade volumes — could reveal how political shocks map to debt and inflation cycles. Global commodity price histories (grain, coal, oil) would complement the wages and CPI series to study real purchasing power across centuries. For modern validation, pairing with ONS (Office for National Statistics) post-2016 series would extend coverage and test whether long-run relationships hold in the contemporary period. Finally, climate and harvest records for the medieval and early modern period could open rare correlations between agricultural shocks and early GDP or price data dating back to the Domesday-era entries.

Columns

year (date)
variable_id (string)
variable (string)
section (string)
unit (string)
value (decimal)