How These Numbers Are Built

Every figure on Legal Cost by State comes from a table researched state by state, and every table follows the same rules. This page sets out those rules, the calendar on which each table is checked again, and what each published dataset covers. The counts are computed from the data each time the site is built, so they describe the tables exactly as published.

The rules every dataset follows

Sources stored with the values

Each value is stored with the address of the page it came from: a statute, a court rule, a court fee schedule, a government table or a national comparison table. The dataset keeps those addresses value by value, and the chart shows a source link on every row. Where a dataset holds values with no source of their own, its section below gives the count and the columns.

Primary and secondary sources

A primary source is the law itself or the body that applies it: a legislature's statute text, a court rule, a court self-help page, a court or clerk fee schedule, or a government table. A secondary source summarizes the law: a national comparison table, a legal aid page or a law office's published list. Each value is labeled one or the other, either by the researcher in a source-type column of the dataset or from the source's web address, where legislature, court and government addresses count as primary. Some datasets add a third label, statute mirror, for statute text read on a site that republishes it because the official site could not be read.

Blanks stay blank

When the research found no reliable figure, the cell is left empty and the table says "Not published". Nothing is estimated, averaged or borrowed from a neighboring state to fill it. A blank means the figure was not found in that pass, not that the rule does not exist.

Values marked for a check stay out of headlines

Some values rest on a copy, a summary or a partial source because the official page could not be read: a filing fee taken from a national table, one county's schedule or a saved copy of a court page, or statute text read on a site that republishes it. Those values stay in the table with a marker, such as "court check pending" on a divorce filing fee or "(statute mirror)" and "(secondary index)" on a probate cell. They are left out of every headline figure (the lowest and highest values and the counts in a page's quick answer) until they are confirmed against the official source, because a headline figure is the one most likely to be quoted on its own.

Checked-on dates

Every dataset records the date its rows were last checked against their sources, each chart prints that date above its table, and the dataset keeps it beside the rows. The date at the top of this page moves whenever a dataset below changes.

What the site never publishes

A lawyer's fee appears only where a published source states it (a statute, a court order, a published survey or a provider's published price list) or as arithmetic on a fee schedule written into a statute, labeled as the site's arithmetic, and the source is named. The site does not publish fee estimates of its own, averages of quoted prices or fee ranges without a published source. Where no source publishes a fee, the page says so. The tables state rules and published figures; they do not assess anyone's situation, and the site does not recommend, rank or evaluate any lawyer or firm.

Refresh calendar

Each data project has a row in the refresh calendar: what is checked, how often, the last recorded refresh and the next date it is due. A row whose due date has passed must be refreshed, or the delay recorded, before the site is updated again.

What each dataset covers

One section per published dataset, 8 datasets in all. A value is one cell of the data: a published value holds a figure or a rule, and a blank is a cell the research left empty.