Counted vs estimated road traffic figures

Labels describe production method

Counted means a roadside count informed that location’s figure for stated year. Estimated means DfT calculated value through another approved method, such as applying growth from earlier data, using an adjacent link or deriving flow from surrounding evidence. Estimated does not mean invented. Counted does not mean complete 24-hour observation on every day.

Both values are AADF estimates. Even counted AADF usually expands a limited observation into an annual average using traffic profiles. Difference lies in how directly current local observation contributes.

Why every road cannot be counted every year

Great Britain contains far more road links than can be surveyed annually. Major-road links follow rolling count cycles. Some may be counted every two, four, eight or—on parts of Scottish and Welsh networks—twelve years. Minor roads use samples because counting every local street would be impractical. Annual statistics therefore combine fresh observation with modelled continuity.

Common estimation routes

• Grown: earlier AADF adjusted using observed traffic change for comparable roads.
• Dependent: value linked to adjacent road section, often where administrative boundaries split one continuous link.
• Derived: surrounding links used where direct counting is unsuitable, including very short or unsafe links.
• Automatic counter: data supplied from continuous equipment and converted into AADF.
• Previous-year based: earlier link estimate carried forward using stated method.

Exact wording shown under “Data quality” on each result matters more than simple badge alone.

Practical confidence ladder

Fresh counted figure at exact location offers strongest site-specific evidence in this dataset. Recent estimate with recent underlying count is usually useful for broad comparison. Estimate based on much older local observation deserves more caution, especially where development, road layout or travel behaviour changed. Old minor-road sample point may describe history better than present conditions.

No universal age cut-off turns data from good to bad. Stable rural link may change slowly; street beside new housing or bypass may change sharply. Context controls risk.

How to compare without fooling yourself

1. Prefer same year.
2. Prefer same method where possible.
3. Check last actual count year.
4. Compare similar road types.
5. Look for abrupt historical jumps.
6. Read junction description and map position.
7. State uncertainty when publishing result.

Do not rank two streets to nearest vehicle when one value is fresh counted data and another is carried forward. Better conclusion might be “Road A appears busier in available DfT estimates, but Road B’s figure relies on older evidence.”

When to seek another source

Planning applications, noise disputes, school-street design, freight routing and property decisions may need current peak-period surveys. Local highway authorities, planning documents or commissioned automatic counts can add time-of-day and directional detail absent from annual link estimate.