GAMSO discussion - January 28th, 2025

[Chris-ECE]

General Approach to the Revision of GAMSO

The 2023 consultation feedback related to GAMSO has been clustered according to area of GAMSO in the google doc. Comments not relating to a particular activity area are clustered according to the topic of those comments, which are sometimes cross-cutting.

Some countries said they were happy with the current version of GAMSO, or even asked us not to change it (section 1.1), while others asked for more detail (sect 1.2). InKyung pointed out that adding more detail would overlap with the Handbook on Management and Organization of National Statistical Systems, which are already somewhat influenced by GAMSO and GSBPM:
Would could instead link to the Handbook from GAMSO, as a resource that is based on GAMSO.

Gabriel pointed out that GAMSO is less mature than GSBPM. He was not sure they would get GAMSO to the same level of structure as GSBPM. He suggested rather than a new version of GAMSO should reflect the trend that that the role of NSO is expanding to data governance and data stewardship, leading to questions about interoperability of data so that it can be reused for statistical purposes.

Joni said that GAMSO is clean and tidy, but not built to the same standard as GSBPM. He suggested considering the long term strategy of what to do with GAMSO: We could move towards helping to align it with GSBPM to be merged in future. We can try to bring up the quality of the document (numbering, the way that it reads, etc). He said numbering is helpful when using enterprise architecture software, to sort items. However, others suggested avoiding numbering, based on similar discussions about this within the GSBPM revision.

Edgardo suggested not merging GSBPM and GAMSO, because they are at different levels, by which he means that while there is a hierarchy, it is not enough for them to be merged into a single model. He also said that the way that GAMSO refers to activities makes it like a reserved term, which makes it hard to reconcile the word “activity” with other models such as GSBPM: In statistics, most activities are related to production processes rather than administrative activities. (Chris mentioned that the new version of GSBPM, we’ve introduced the idea that GSBPM refers to (process) activities.)

Chris pointed out that GSBPM refers to a single instance of a production process, whereas GAMSO refers to multiple production process instances, and asked whether this is an inconsistency.

Juan said GAMSO provides a context from where statistical processes are being driven. An activity can be performed upon many instances of a statistical production process. He said there are inconsistencies in the way we talk about activities and processes. With GSBPM we are zooming in to focus on one special activity (production) because of its special concern to statisticians. But the other (non-production) activities also have their own processes, which may relate to production processes. For example Human Resources are relevant to production. – We can start by having an introduction that makes this connection clearer, and consider defining activities more precisely (in more detail). When you implement GSBPM, you also need GAMSO also for the transversal activities.

InKyung ask what’s the target audience and main purpose of GAMSO? She suggested that GAMSO is mainly for middle-managers rather than for more technical staff like software engineers, so might not benefit from having extra detail.

Joni said he has used GAMSO to understand organizational design, and costings for different activities, or measuring productivity and investment (e.g. how are you investing in new methodologies and managing new methods). For some software projects, it has also been a useful framework for understanding ancillary activities that go alongside production. For organizational design and transformation programmes, you want to look at the whole organization in addition to how different types of statistics are collected. He would like Figure 2 to contain the GSBPM.

InKyung said that when considering new technologies like Cloud or AI, people may look at GAMSO to see which parts of their organization might be impacted. Some organizations don’t have centralized support for methodology or quality, so they GAMSO provides an idea for which aspects could be centralised in their organization, and to identify core functions for possible restructuring of their organization (almost like a blueprint).

Edgardo said that GAMSO is about how you organize the work, rather than a detail of activities such as in GSBPM, which now goes down to the level of Tasks (which might have been called activities). He said the name of the model should reflect that it’s focused not so much on activities, but rather on organising statistical offices, or more specifically about organizing production: It could be called Generic Organization Model for Statistical Production.

Joni suggested not to mandate an organizational structure, and said that GAMSO activities will take place regardless of whether there exists a part of that organization dedicated to it. – Edgardo clarified that he was not suggesting to specify an organization’s structure, but specifying a typology of things (activities). While GAMSO describes “activity areas” at the highest level, the lower level refers to singular activities, though they are in fact also groups of activities.

Gabriel suggested thinking about the coherence and interoperability between GAMSO and GSBPM and perhaps in the future to see if there’s an opportunity to unify them. He said for the structures shown in the graphics, Strategy and Leadership seems to be at the same level as Capability Development and Corporate Support. He suggested depicting Capability Development under Strategy and Leadership, and Corporate Support under Capability Development. This is inspired by a management (hierarchical) view of the organization. Chris suggested a pyramid depiction.

InKyung mentioned that GAMSO doesn’t have a description of how different activity areas inter-relate. We could have a section describing this. Juan said that the utility of explaining how these activities inter-related is in how they enable information production.
Action: Juan offered to create some picture to illustrate this point.

Regarding the suggestion about stacking the different activity areas according to management hierarchy, Edgardo wasn’t sure that Capability Development activities would necessarily impact the Corporate Support level below it. So this could be misleading if illustrated this way.

Chris: Should the next call focus on the challenges and changes facing NSOs? (Cloud, data stewardship/governance, using third-party data, etc)

Gabriel suggested that we think a little further about this to decide what to cover in the next call: What do we want to achieve with GAMSO? When did we use GAMSO, and for what? InKyung agreed, but we have a time pressure, if we agree GAMSO is a blueprint for statistical organizations to know what activities they should engage in, a priority could be to identify what activity areas should be added to that it doesn’t have, like data governance/stewardship.

Gabriel suggested that Canada, Australia, New Zealand and others in Europe have very strong data governance roles, and might be able to provide insights. Also, having organizations that integrate geospatial (such as Mexico) might be useful.

Action: Jimena Juarez (who is in the office of the Chief Data Officer at INEGI) offered to talk to colleague to see if they have advice to share on this.