Derived measures — location quotient and shift share

Two standard measures of regional economic analysis are computed from DTG industry data: location quotient and shift share. Location quotient is available today. Shift share is not yet available: it is specified, and it is not in the app, the API or an export. Both are calculated at query time from the underlying employment values rather than stored, so both always reflect the data in the vintage being queried.

Neither is a new measurement. Each is an arrangement of employment figures that makes a particular comparison easier to see.

Location quotient

A location quotient compares how concentrated an industry is in an area against how concentrated it is nationally.

It is the industry’s share of employment in the area, divided by that industry’s share of employment nationally. A value of 1 means the industry accounts for the same share of local employment as it does of national employment. Above 1 means the industry is more concentrated locally than in the country as a whole; below 1 means less.

DTG computes location quotients on employment only, against the nation as the reference area, at every industry level and for every geography, including area clusters.

Reading it

A location quotient describes concentration at a point in time. It says nothing about growth, about whether an industry is thriving, or about why the concentration exists. A declining industry that has always been locally dominant will still show a high quotient.

Concentration is also not the same as size. A small county with one large employer in a narrow industry can produce a very high quotient on a small number of jobs. Read the quotient alongside the employment count it came from.

Not yet available. Nothing on this page about shift share describes a feature you can use today; it describes the method that will be used.

Shift share decomposes the change in an industry’s local employment over a period into three parts, each answering a different question about why the change happened.

National growth is the change the local industry would have seen if it had grown at the rate of the national economy as a whole. It reflects the general economic tide rather than anything about the industry or the area.

Industry mix is the additional change attributable to that industry growing faster or slower than the overall national economy. It reflects the industry’s national trajectory, not local conditions.

Competitive share is what remains: the difference between the change the area actually saw and the change implied by the national rate for that same industry. It is the part that national trends do not explain, and it is generally the component of most interest.

The three components sum exactly to the observed change in local employment. Each is calculated against the area’s employment at the start of the period, and the nation is always the reference.

Periods

Start and end periods are selected by the user, on either a quarterly or an annual basis, across any periods available in the vintage.

This includes projected periods. A shift share that ends in a projected period decomposes projected change rather than observed change, and should be read as a decomposition of the projection, not as a measurement of what happened.

Reading it

Competitive share identifies where an area is outperforming or underperforming the national trend for an industry. It does not say why. A strong competitive share is a signal to investigate local conditions, not a finding about them.

The components can also point in opposite directions, and that is often the useful result: an industry can add jobs locally while showing a negative competitive share, because it added fewer than its national growth implied. Raw job change alone would not show that.

Provenance of derived measures

Neither measure carries a disclosure code or a reliability score of its own.

Both are computed from underlying employment values, and their accuracy depends entirely on the accuracy of those values. Where the components are published QCEW figures, the derived measure is as solid as the data beneath it. Where they are estimated, it inherits that estimation.

This matters most in small areas and at detailed industry levels, where suppression is heaviest. A location quotient built from two estimated values is itself an estimate, and a shift share across a period where the endpoints are estimated carries that uncertainty into every component. Before relying on either measure in a small area, check the disclosure codes and reliability scores on the employment values it was built from. Both are described on the disclosure and provenance page.