The ASHE data is provided as Excel files, needing Microsoft software to access it. It is also tricky to find the right data sheet, having to navigate through a number of HTML and PDF documents.
However the path to the Excel data seems well-structured so I though I would try to provide an interface and converter to XML, using the http:/elev.at to do the heavy lifting.
The structure of the path is defined in an XML file from which the interface is generated by an XQuery script. This is generic so it could be used on other datasets, all but the sheet number. This has to be custom-mapped from the gender and employment selectors.
This has hardly been tested and the structure has been infered from browsing the site. There may be variations which I didn't notice (The change in the format from 2008 to 2009 - underscore to hyphen - was tricky)
All tables are not present in all years so there will be some bad links.
Occupation and Industry categorisation has changed during the time period. Other classifications and methods of collection may have too. The ONS site provides the detail.
Data may be lost in the conversion to XML, particularly the relationship between footnotes and data. Data on the reliability of data, contrary to my earlier posting, is provided in full by the parallel Coefficient of Variation datasets - my apologies to the ONS.