Collective strength

It's National Libraries Day. I could  stand at the railings outside what this time last year was my local library, pining for services lost but I'd much rather think of positive things.

National Library Day is the date chosen for the public launch of the Greater Manchester Libraries Consortium.

The consortium's been a work in the wings for the best part of five years, three years in concrete form. The heart of the technical architecture is the library management system: we started sharing the bibliographic data three years ago and this first phase of the shared lending functions was technically live just before Christmas. It's a small but important step: customers can look at an online catalogue shared by the whole consortium — Blackburn, Bolton, Manchester, Oldham, Rochdale, Salford, Stockport and Trafford — and reserve an item from any of them.

At this stage you'll need to go to the item's "home" library to pick it up, there's a big piece of work to be done on transit logistics across the consortium but even with this caveat it's an exciting first step and it was good to see not only a big piece in the Manchester Evening News but also a very approving editorial. We don't often get high-profile hugs these days, we should enjoy them when they come along.

Strangely enough, although reservations are easily the most complicated part of the lending library operation this phase has been the easiest to implement because for all intents and purposes each authority's operation is still acting in a stand-alone basis.The next phase will require more of an integration of circulation systems and the scoping work for that is going on at the moment.

Like any decent undertaking like this we've had our fair share of stumbles and oh no moments, and one of the painful consequences of sharing a system across a consortium is the need for a fair bit of coordinating effort to get the ducks in a row ready for routine system upgrades. But there are upsides: we've built up a couple of useful support networks and the benefits have certainly extended beyond the confines of the library management system and the consortium.

And the public have a new extension to their library services and the promise of more to come.

Data sharing between libraries

We're at the stage in the evolution of the AGMA library consortium where we're starting to work through the practical — and legal — implications of shared services.

• Sharing our catalogue data is relatively easy: the data standards are well-established and most the data itself is published in the public domain on library OPAC's, etc. Which doesn't mean that it was all plain sailing and we've not got some more work to do. 

• Sharing borrower data is obviously fraught with all sorts of information governance and data protection issues on top of the problem that there isn't any data standard save that imposed by the structure of our shared LMS and the commonalities we've discussed and agreed on a case-by-case basis.

• Virtually every circulation dataset is a back door into the borrower data.

I've been thinking through some of the questions we need to be asking ourselves on this journey. It's still early days so isn't exhaustive; at this stage I'm trying to work out what we need to worry about at a general level prior to starting work on a risk analysis.

Purpose	Type of Information	Recipients	Data Controller	Notes/queries
Membership information including contact details –voluntary service, customers will be asked if they want to opt in	Customer name, address and contact information, DOB. Disability, ethnicity and other demographic details Family relationship details Lending history	Library staff (including all other authorised Spydus users) of approved Authorities within the scheme	Local Authority (Data Subject’s Local Authority will be the data controller)	Which data is to be shared? Is it all or nothing? If partial, which parts and how managed? Same question applies to who the data is being shared with What would be the position of volunteer-managed community libraries? How do we switch sharing on/off? What happens if a customer changes their mind? How are they “quarantined?” What happens to the data held in loans, charges and reservations? What happens to any outstanding loans, fines and charges? Who owns (and is responsible for) the data?
Loans information	Details of the loan including borrower, item, location and status of loan. Loans history	Library staff Specific customers can see all details of their loan(s) All customers can see some details of the loan(s)	Local Authority (which?)	This is the crucial element to be managed: It is the purpose of the data-sharing agreement It is the bridging element between the personal customer data and nearly all the other data sets There is a hierarchy of viewing permissions If a customer has said “no” to data-sharing, how is the borrower data in the loan, charges and reservation records expressed? If the customer changes their mind about sharing their data, is it automatically redacted from these records? Who owns (and is responsible for) this data? Whose loan policies? Applied from the lending library? Including fines and charges? How do exceptions apply? “Non-default” borrower types and collections
Overdue/pre-overdue notices	Contact details including borrower name, address, telephone and email; loan due dates and items involved	Library staff Specific customer	Local Authority (which?)	Derived from loans data and subject to same questions It would make sense to aggregate these to improve efficiency and save costs (see notes on charges, etc.)
Reservations	Contact details including borrower name, address, telephone and email and items requested	Library staff Specific customer	Local Authority (which?)	All the questions for loans apply for reservations (which are effectively loans-in-waiting) Whose charge régime applies? Would the Data Controller be the “owner” of the customer record, the library that placed the reservation or the library it will be picked up from (if a different library authority)?
Requests	Contact details including borrower name, address, telephone and email and items/articles requested	Library staff Specific customer ILL system (bibliographic and/or article data only)	Local Authority (which?)	In nearly all respects as reservations, just more complicated charges [The operating procedures would probably need modifying in the light of the shared lending environment.] This will need to be revised in the event of a fuller integration with UnityWeb or equivalent third-party systems
Notifications for any reserved items	Contact details including borrower name, address, telephone and email and items requested	Library staff Specific customer	Local Authority (which?)	Derived from reservations/requests data and subject to the same questions It would make sense to aggregate these to improve efficiency and save costs (see notes on charges, etc.)
Charges/fines/fees	Contact details including borrower name, address, telephone and email; details of the transaction that generated the charge	Library staff Specific customer	Local Authority (which?)	Derived from loans and reservations/requests data and subject to the same questions How will these be managed: Payable only where incurred? Payable globally? Impact on traps/alerts (whose parameters apply?) In the event of recovery, who legally owns the charge? In the light of the above, what would be the effect (if any) of aggregated notices?
Catalogue/ discovery records — bibliographic data	Title-level catalogue data	Library staff Library customers and general public	Local Authority (which?)	Bibliographic data – already shared data Don’t forget that there is a link to the borrower record from the review/rating in the bib data in Staff Enquiry Potentially links to more than one Data Subject, so which would be the Data Controller for this catalogue data? Shared responsibility? How? Similar questions are required of other customer-created content such as tags (these are lost in the current versions of Spydus 9) (Not all data are published for the public)
Catalogue/discovery records — holdings/item-level data	Catalogue data, including electronic holdings	Library staff Library customers and general public	Local Authority (Which?)	Holdings data Links to personal data via loans/loan history and status/status history Potentially these link to more than one Data Subject, so which would be this Data Controller for the catalogue data? Logically should be the owner of the holding item (Not all data are published for the public)
Management Information/ Business Intelligence	Reports detailing usage of service, per location	Library Managers	Local Authority (Data Subject’s Local Authority will be the data controller)	Essentially should be summary data, though we’d need to have safeguards against breaches caused by very small sample data Proper safeguards and risk analyses are required before making this data available to third parties
	Demographic breakdowns	Library Managers Designated authorised analysts	Local Authority (Data Subject’s Local Authority will be the data controller)	Most would be summary data, though we’d need to have safeguards against breaches caused by very small sample data Some data (e.g. lists of postcodes) are granular enough to easily identify Data Subjects so safeguards need to be in place on the use and presentation of this data are required before making this data available to third parties
	Marketing databases	Library Managers Designated authorised marketing staff	Local Authority (Data Subject’s Local Authority will be the data controller)	Is the “I agree to receive marketing” (or equivalent) field global or local? The selection of data explicitly must be limited to those customers who have agreed to contact so as to comply with Privacy and Electronic Communications Regulations. Proper safeguards and risk analyses are required before making this data available to third parties
	Stock management data	Library staff Designated authorised third-party service providers	Local Authority (which?)	Nothing pertaining to Data Subjects should be included in this data. Stock ownership should be straightforward. Stock usage more problematic: Global usage figures recorded against bibliographic/holdings data? Local usage only? How would (if at all?) third-party stock analysis systems like CollectionHQ differentiate between local and extralimital use? In the early days at least there will be pressure to be able to provide evidence that stock is being used “fairly” with local library customers having first dibs for local stock
	Ad hoc data requests	Library Managers Designated authorised third parties	Local Authority (Data Subject’s Local Authority will be the data controller)	Most would be summary data, though we’d need to have safeguards against breaches caused by very small sample data Some data (e.g. lists of postcodes) are granular enough to easily identify Data Subjects so safeguards need to be in place on the use and presentation of this data Proper safeguards and risk analyses are required before making this data available to third parties FoI requests would be subject to the proper exclusions
SIP2 data	Data used for interfacing between Spydus and third-party systems	Library staff Specific customer	Local Authority (Data Subject’s Local Authority will be the data controller)	The particular case at the moment would be where data held in the customer record determines the access or not to third-party systems and services. Would the data be determined globally or locally? Standard use of data fields? Standard coding sets?

I'd be interested to know if/how this analysis sits with the experience of established consortium libraries, especially if I've missed something that could cause us problems.