Summary

• The ICO has launched a consultation on how it plans to prioritise Freedom of Information (FOI) complaints.
• This document is mySociety/WhatDoTheyKnow's draft response and thinking, which we will revise based on feedback from users and other interested parties.
• The ICO thinking strategically about how to prioritise case work is a really useful exercise; our main concern is whether there are better, higher impact uses of time than those proposed.
• The purpose of the complaints system is not to release information, but to steer the whole FOI regime. Rather than focus on high impact requests, it would be better to look for strategic interventions to unblock common problems experienced by high impact requesters.
  • Recommendation: Add a prioritisation criteria and process around administrative silence/stonewalling.
  • Recommendation: Use discretion to pay special attention to complaints from authorities where there are repeated problems.
  • Recommendation: Build statistical knowledge of the whole FOI system (especially internal review) to support earlier and more effective intervention.
• We disagree in principle with prioritising classes of requesters (eg. journalists and civil society organisations), and on practical grounds do not think this is the most effective way of achieving the ICO's goals.
  • Recommendation: reframe these as criteria for the kinds of requests from these groups that would benefit from faster treatment (eg time sensitive requests, stonewalling).
  • Recommendation: Alternatively, redefine criteria as kinds of activity rather than requester (‘journalism' rather than ‘journalists'). Make these more porous categories (where many requests may qualify) and only fast-track a certain percentage of them to manage the overall volume.
  • Recommendation: As one of these porous categories, if the destination of the request is to a public repository (like WhatDoTheyKnow), this should be seen as a request having a higher public impact.
• The consultation reflects a view that the impact of FOI is primarily through intermediaries, making high impact requests in the public interest. We believe that the public interest needs to be viewed as wider than the result of high impact requests. There is enormous public value collectively even for privately motivated requests.
  • Recommendation: Change the definition of frivolous from ‘low public interest in information requested' to ‘low public interest in pursuing the complaint'.

Introduction and context

This is a draft response to the ICO's consultation on how it prioritises Freedom of Information complaints.

We are releasing this draft before the deadline for feedback so we have time to get input from WhatDoTheyKnow users or other interested parties.

This submission was revised based on feedback from users and talking to other organisations (see final section for the final submission).

If you'd like to make your own response, you can also respond to the ICO directly. You do not have to have an opinion on all questions to make a response.

We have bullet point summaries of our position on each question at the bottom of this document.

The deadline for submissions is 5pm on 19th December 2022.

About mySociety

mySociety is a not-for-profit group, based in the UK but working with partners internationally. We build and share digital technologies that help people be active citizens, across the four practice areas of Democracy, Transparency, Community and Climate.

mySociety runs WhatDoTheyKnow.com in the UK, and maintains the Alaveteli software and network which powers Freedom of Information sites around the world. This software allows users to make public requests through a website that are sent to authorities via email.

mySociety produces research into civic tech and systems of access to information, and has published analysis (and proposals for reform) of Freedom of Information in the UK and across Europe.

About WhatDoTheyKnow

WhatDoTheyKnow.com is a website where people can make Freedom of Information requests online using a form. These requests are sent via email to the appropriate public authorities.

The requests and responses are also displayed in public, meaning that information is available to all users of the site, rather than just the original requesters. WhatDoTheyKnow makes Freedom of Information more efficient, because information requested once is accessible to a wider range of people and discoverable through search engines. This reduces the need for different people to request the same information, and means that information may be found and used by people who would never make a request themselves.

The moderation of this public archive is managed by a small team of dedicated volunteers with the support of mySociety's staff and board.

What the ICO wants to do

The ICO is consulting on a change to how it processes FOI/EIR complaints. They want to end the current "cab rank" principle of case assignment (first come, first served), and create a fast-track system that accelerates the timetable for information being requested in the public interest.

In short, they want to free up resources by closing some kinds of appeals faster and with less effort, and spend the freed up time on higher impact cases.

What currently happens

• After internal review of an FOI decision (or in the event of no reply at all), a requester can appeal to the ICO for an independent review of the decision. These appeals enter a queue, and are eventually assigned to a caseworker.

How the ICO is planning to free up time

• Issue less detailed decisions (especially when upholding the authority's original decision).
• Spend less time writing decisions that do not meet prioritisation criteria and spend more time on those that do.
• Be explicit that they will not accept complaints more than 6 weeks after a decision has been received by the requester (except under exceptional mitigating circumstances)
• Tighten policy on not engaging with frivolous or vexatious complaints:
  • Redefine ‘frivolous' - from "when a complaint has no serious intent or is considered unworthy of serious treatment" to "when there is such a low public interest in the information requested that it would be a disproportionate use of our resources to investigate it".
  • Take the view that if the public body has followed guidance on judging vexatiousness, they will not make a decision (as the complaint itself is then vexatious).

How the ICO proposes to spend this time

• Based on a set of criteria, apply a fast-track approach to some public interest requests so that they will be assigned to a case worker sooner, and generally proceed through all steps more quickly.
  • It is worth being clear that the proposed process is faster, but not necessarily considerably so - the goal is that the case will be allocated within four weeks rather than six (proposed KPI).
• Where possible, the ICO will make decisions quickly based on information available when received.
• When engaging with public authorities on prioritised cases, the ICO will set short deadlines, on the basis that all information and documents should be readily available after internal review.
• The ICO will use information notices to compel a response when public authorities do not meet those deadlines.

The ICOs have proposed criteria for their public interest track. Requests that trigger any of these criteria get routed to the fast stream system:

• High public interest in information
  • Significant media interest
  • Request involves a large amount of public expenditure
  • Information requested is needed to respond to significant public consultation
• Class of requester
  • Journalists
  • Civil society group (or otherwise working on behalf of others)
  • Elected representatives.
• Vulnerable groups or people affected by information
  • Information covers policies, events which have a significant impact
  • Information has a high potential impact or harm on people nationally or in a particular court case.
  • Information can directly affect requester's health (impact on treatment, live court case)
• Operational reasons
  • Novel problems - basis for wider guidance
  • High impact in clarifying one case. For instance, several similar requests/part of round robin.

Key context: ICO funding

The key context for this debate is the highly constrained funding of the ICO. While the number of FOI requests has increased over the last 15 years, and the corresponding complaint caseload has increased, the ICO's budget has not.

Our view is that the ICO is under-funded, and investment in the FOI regulator is an extremely cheap way of unlocking the wider benefits of an access to information system. Our research on FOI across Europe has a summary of the evidence of an effective FOI oversight body in reducing corruption and improving governance outcomes.

But that is not what this consultation is about, and the ICO consultation is explicit about the trade off. Acknowledging that "at the end of every complaint is a person simply trying to exercise their statutory right to get hold of information they need", they argue that they "need to make better choices about how we allocate our resources to those issues with the highest impact".

There is no easy way out of this dilemma. The current approach offers procedural fairness to complainers, but leads to long delays (and a significant backlog) that undermines the wider FOI system. But while some form of prioritisation can be justified, we have concerns about the specific criteria the ICO has chosen, and the way it has framed the wider impact of Freedom of Information.

The public interest and FOI

In the ICO's consultation, the public interest criteria and definition of frivolous complaints reflect a view that high impact requests in the public interest happen mostly through intermediaries. This is in part the justification for highlighting specific groups like journalists as potential candidates for fast-track attention.

This is an understandable belief, given the greater visibility of these groups of requesters, both in terms of the outputs of FOI requests, and in public debate about FOI. Giving evidence to Parliament, the previous Information Commissioner said that "one in 1,000 citizens in the UK will file a Freedom of Information Act request, but journalists are standing in their shoes. It is through journalists that the public can understand or get to know why decisions are being made on their behalf. Journalists, public interest researchers and advocacy groups are important requesters".

The problem with this view is that, based on our polling and the ICO's own polling, this figure is 100 times too small. Closer to 1 in 10 people have made a Freedom of Information request. Freedom of Information is not a niche right, or mainly used by journalists, but has been used by millions of people. Most use of FOI in the UK is completely outside official statistics, which distorts our understanding of who is benefiting from the system.

Given this, the ICO needs a wider understanding of the public interest. Most FOI requests have at least the potential for interest by someone else. This is the basic premise of WhatDoTheyKnow, where FOI requests are published in public, and made more accessible through internet searches (leading to millions of views of the collection every year).

But even requests that do not fit into that category are worth defending. Citizens should be able to access information from the public institutions that have power over their lives. They should be able to do this, even if it benefits no one but themselves - because the existence of that right shifts the behaviour of authorities in a way that benefits more than each individual requester. It is in the public interest that the FOI process is effective even for entirely private requests.

This highlights a problem with the proposed new definition of frivolous complainants. This definition focuses on there being "such a low public interest in the information requested that it would be a disproportionate use of our resources to investigate it". Many individual requests may not have a public interest in the information collected, but collectively they have an important impact which shouldn't be defined out of the ICO's interest.

A modification about the "public interest in pursuing the complaint" reflects the spirit that the Commissioner has expertise in recognising complaints that, practically, are going nowhere, while recognising the ICO's role includes defending FOIs made by private interests.

Maintaining the system

The primary purpose of the ICO appeal process is not to release information, but to uphold the much wider system of Freedom of Information itself. It is impossible for the ICO to be well resourced enough to release all the public interest information through appeal. If the system is to work, almost all of it has to be released without prompting by the information commissioner. The ICO's role is to ensure the incentives are in place that this happens by default.

These incentives can include a prioritisation system that says some complaints are more important to the health of the system than others. A cab rank principle treats all complainants the same, but a better system for all requesters (and ultimately the beneficiaries of information released) might be to strategically prioritise certain kinds of problems or categories of requests.

The ICO's proposed fast-track categories are focused on the public benefit of the information released, or operational reasons for prioritising a request. This misses potential categories of complaint where prioritising might improve the effectiveness of the whole system.

Stonewalling

Stonewalling/administrative silence is where, rather than explicitly rejecting a request for information, the authority simply does not answer. Due to an oversight in the law's construction (but addressed in Scottish FOI law) the request enters a state of limbo where the lack of response cannot be appealed. The only solution is to complain to the ICO, who can compel a response through a Decision Notice. However, the timeline on responding to a decision notice (35 days¹), is longer than the 20 day window during which they should have replied in the first place. Going through this process adds considerable delay to the response.

In the absence of legislative reform, the ICO can use its prerogative to prioritise certain kinds of complaints to make this practice less effective to delay requests. This is a kind of complaint that by definition does not engage complicated questions about exclusions, and could be addressable through a streamlined process on a faster timeline. Some parts of the timescale are unavoidable, but the ICO has made efforts recently to make some aspects faster. If more resources were dedicated to this process, it could go further.

The ICO has already improved the time between when it receives a complaint and when it sends its initial hurry-up message to the authority, in recent requests taking only a few days before contacting the authority and asking them to respond to the original request within 10 days. The ICO's form letter to follow-up with an authority who hasn't responded itself says the authority has 10 days before it will be reported to the ICO. In combination all these soft warnings add up to significant delays

If the soft warning approach does not work, a significant amount of time can pass before the decision notice is sent. In DN IC-194023-H8Q3, 36 days pass between the complainant notifying the ICO, and a decision notice to release the information being issued, in IC-190744-J2F9 57 days elapsed. In both cases, the ICO acted within days to contact the authority and encourage a response, but there is then a substantial delay before the 35 day clock is triggered.

This feels like a process where some special attention and resources could have significant rewards. The ICO could additionally indicate it would respond speedily to follow-on complaints about the outcome as a way of discouraging further delays. The ICO could also incorporate previously having had to issue a decision notice in this area to that authority as a reason to use a faster process.

Special measures

The ICO is proposing giving some kinds of complainants different treatment, but this could also apply the other way around - with faster attention given to certain authorities. This could be used to construct ‘special measures' sanctions, within the present legislation.

Where authorities are consistently non-compliant (or acting within the letter of the law, but breaking the spirit), the ICO could adopt a policy of paying close attention. As part of its planned accelerated process, the ICO has already indicated it can reasonably anticipate faster responses from authorities to requests from the ICO (especially when the information should have been gathered either initially or as part of internal review). The ICO could maintain a list of authorities that complaints against will be reviewed speedily, with expectations of faster response times from that authority. This could, for instance, involve accepting complaints about a non-response to internal review after 20 days rather than 40 days.² Administratively, the ICO could ask requesters to inform them of follow-up requests for internal review and generally take an interest in this kind of delayed request.

Another approach would be to shift resources from speeding up individual reports, towards the ICO's other legal powers. For instance, issuing practice recommendations for organisations with high numbers of stonewalling complaints. 2020 analysis by OpenDemocracy³ found a small number of central government departments were responsible for the majority of stonewalling complaints to the ICO. A targeted use of resources could help improve how outlier authorities respond, and have impact far beyond the initial complaints.

Clear responses to authorities gaming the system are not just useful in releasing information from that authority, but are part of encouraging all authorities to act in accordance with law and code of practice. A focused application of resources on the worst offenders would be effective at ensuring the release of more information in the public interest. This would also express a principle of fairness to all authorities, that those who follow the rules are given greater respect for their autonomy.

Avoiding a two tier system

The controversial aspect of the ICO's proposal is to prioritise based not just on the content of the request, but on the kind of requester. Their list covers journalists, civil society groups, and elected representatives.

There are reasonable motives to prioritise specific classes of requesters. Taking the example of journalists, they may be a small part of the overall volume of requests, but have the potential to produce especially high impact requests. Delaying tactics are especially effective on journalists - obstructive authorities may successfully delay the release of information past the point where the information is newsworthy. FOI is supposed to be requester-blind (decisions are not made based on who the requester is, but what information has been requested), but in practice requests from journalists are identified and sometimes treated differently. The regulator bending this rule itself as a corrective might be justified.

Against this are practical questions of what is the best use of the ICO's limited resources to help groups like journalists. Our view is that this extra category does not achieve much that is not already covered by other criteria and raises practical questions around how exactly these classes are defined and enforced, as well as the risk that these broad categories will overwhelm the fast lane approach.

Requester-blind principle

FOI is a general right of citizens, not a special mass media law. A fast lane for journalists (or any special group) is leading away from the spirit of the law. The ICO stresses that this prioritisation would not affect the outcome of the case, which would be judged on the facts. They argue that the process remains requester-blind - as the decision process remains on the question of whether the information should be made available to everyone, rather than any consideration of the specific person. But prioritising requests made by some groups, while effectively de-prioritising other kinds of requests made by others, raises issues that need to be addressed.

In the ICO's proposal, this deprioritisation of requests by ordinary complainants happens in two key ways: expanding the definition of frivolous requests to include those where there is low public value in that specific information being released; and reducing the time spent on drafting replies to requests (eg by not recapping the entire case). Neither of these inherently changes outcomes, but there is an explicit trade-off made between this and more time spent on the detailed descriptions of prioritised cases.

This is a trade-off that might be justified on the merits of the case (spend less time on uncontroversial decisions, more time on decisions that may need to be justified on appeal) - but is much more difficult where the trade-off is about who the requester is. A requester might be a journalist, but their complaint may be noncontroversial and easily dismissed. That would be a prioritisation slot misspent.

Definitions and scope

A practical problem for special access for journalists and civil society groups is definitions. There is a sliding scale between "obvious established newspapers", through online journalism websites and networks, through FOI requests made by newsletter writers/bloggers and freelancers. Similarly, many grassroots organisations are not registered charities.

The ICO can pick a standard on this but administering any kind of system would have overhead. The US FOIA has a fairly broad definition of "representative of the news media", but a more restricted category of "education or non commercial scientific requesters" that is different from the proposed treatment of civil society groups. Overly restrictive definitions lead to situations such as the German FOI website FragDenStaat publishing a print edition to successfully qualify as press.

An alternative that meets the spirit of both the law and the ICO's intentions reframes these categories in terms of activity. For instance, using a broad definition of ‘journalism', rather than who counts a ‘journalist' is more useful as a wider functional test. The US definition is useful along these lines, similarly in an environment without state licensing of journalists. Ideally, journalists and civil society groups should be able to frame their complaints in a way that triggers public interest tests as a result of their activity without being explicitly mentioned as a group.

Another angle on this would be to focus on the elements of requests by these groups the ICO wants to prioritise. For civil society groups, much of this is covered by the ‘vulnerable groups' criteria. For journalists, adding an additional criteria around time sensitive information could help capture the special vulnerability of journalists to delay.

Importantly, understanding these special cases as activities helps show where requests from members of the public might also fulfil these activities.

Requests by the public, for the public, in public

If it is the destination of the request (and the subsequent impact) that matters, requests made in a way that the response is public on WhatDoTheyKnow (or other similar public archive) speak to potential for impact. Over the last nine years, WhatDoTheyKnow's 660,000 requests have received 170 million page views. The ICO's decision-making process is based on the idea that information is being released to everyone - in practice, WhatDoTheyKnow delivers on this idea.

WhatDoTheyKnow requests can have an impact through non-direct routes. For instance, a request to the Cabinet Office for the Precedent Book via WhatDoTheyKnow was only successful after an appeal to the ICO. The results were highlighted to the group Republic, who in turn framed the information for journalists. This kind of after-the-fact connections of interested citizens, civil society groups and journalists would not be picked up by the proposed prioritisation rule.

Even without an explicit prioritisation based on method of publication many requesters on WhatDoTheyKnow would trigger the proposed ‘on behalf of others' category. Most respondents (54%) to the WhatDoTheyKnow user survey in some way describe themselves as making requests that are relevant to "people in their local community" (43%) or "People in an excluded or marginalised group (people with a disability, LGBT, women, belong to an ethnic minority, homeless, etc)"(18%). If criteria like those proposed were used, WhatDoTheyKnow would be likely to provide templates and language to help users working on behalf of others make more effective appeals.

Sampling from a big net

Having larger categories based around definitions of activity and impact helps prevent arbitrary exclusions and definitions of impact, but may exceed the anticipated capacity of the fast track approach. The solution is to soften the divide between the two tracks by having wide criteria, but using a lottery or random approach to escalate a subset of complaints to meet capacity.

The ICO has estimated that 10-15% of complaints would trigger their proposed criteria. The implications of a wide definition of qualifying complaints mean that the ICO might have underestimated the potential number of qualifying complaints based on the current caseload.

There is also the issue that calculations are based on the current caseload. If successful, a more effective service for journalism purposes would lead to more complaints from journalists than currently arrive. This would be good, as it would reflect complaints discouraged by the current timescales, but also could overwhelm the system, and mean that the fast-track tier eats into resources earmarked for maintaining the current speed of regular complaints.

The fairest way to deal with this is not to artificially restrict the criteria to size, but to have broad categories and then use a lottery to match capacity. For instance, a process of escalating a certain percentage of requests that meet broad criteria is simpler to administer, as it doesn't require fine grained decisions on small differences in public interest, or ruling out groups ahead of time.

It would give a clear control on the fast track, helping meet proposed KPI goals on both tracks and ensure spending more time on high impact complaints, without completely separating the two tracks. In general, tracking referrals under each criteria (and possibly capping some of them) would be useful steps in maintaining control of any new capacity.

Fix systematic problems, not individual problems

If the ICO can free up resources by doing case work faster, it might be true that the highest impact thing it can do with that time is not casework, but being able to better resource enforcement.

It is reasonable for the ICO to reserve some resources and discretion for high priority cases, but stamping down on issues like stonewalling in general, rather than a slightly faster release of a small number of requests, would help make FOI a more practical tool for journalism.

The ICO can't help everyone quickly, and generally intervenes very late in the process. Information of vital importance to public debate will have been requested several months before the complaint can even arrive with the ICO. The most effective way it can spend limited casework resources isn't in releasing individual requests, but by using prioritisation, enforcement powers, and paying attention to authorities as ways of kicking the wider system into faster and more compliant responses.

For the central government, there are useful statistics for helping guide this process, but part of this strand of thinking would be to spend resources building a better statistical understanding of system-wide problems. The ICO (and to be fair, most regulators in Europe) do not have the in-depth statistical knowledge of their counterpart in Scotland. In particular, we are blind to the overall effectiveness of the internal review process (which in other research we've been sceptical of the value of). Developing clearer sight of this process could identify and redress where internal review is used as a delaying tactic (with the ‘real' decision withheld until the requester complaints) and help get useful information into the public domain even faster.

Final response

Feedback

After talking to a few different organisations and getting feedback from some of our users:

• No real pushback on the more technical points.
• General feedback from a few sources that use of discretion is a bit technical and weaker than just reallocating resources for using existing enforcement mechanisms - changing emphasis.
• General support for the idea of seeing requests made in public as having a being component in evaluation of public interest. We're walking a line between "self-interest", but also just a clear statement of what we think is valuable about public archives like us.
• Some pushback on better stats coverage as potentially being an expensive exercise. My sense is to disagree with that - but also we haven't put all our thinking of that in public forms. De-emphasizing a bit in this consultation.
• Interesting suggestion that if there's a fundamental resource problem, a fast track lane could be supported by contributions from the kind of organisational requesters who use it (e.g. journalism). Not going to adopt this, but it does get at some of the tricky issues here around how a blanket ‘journalists' rule does elevate requests from private companies which may (but may not be!) in the public interest.
• One user made a point about the frustration of authorities giving a different line of argument at every appeal - and in favour of the ICO just making decisions based on the information at internal review. The ICO highlighted something of this argument themselves when discussing how they can accelerate certain kinds of requests (and have expressed frustration with this when new arguments appear at tribunal). Will add something stressing that this kind of principle can go further than fast track approaches - and in most cases all the real arguments should be in the open by the end of internal review.

Question 1: Do you agree that, to maximise the benefit from the resources available to the Commissioner for his work on access to information complaints, he should prioritise cases of more significant public interest rather than continuing the ‘cab rank' approach of dealing with cases in date order? If you don't agree, please explain why?

• Yes
• While disagreeing with how public interest is formed in these criteria, we also acknowledge that a cab rank (first come first served) approach is a poor approach to having high impact and managing the overall system.
• Equal treatment of complainers does not necessarily lead to action that is most useful to requesters overall, or of the beneficiaries of released information (including the wider public through intermediaries).

Question 2: Do you agree with the proposed factors that will inform the ICO's decisions on which cases to prioritise? If not, which do you not agree with and why? Are there any additional factors you would include?

• No
• Except for special classes of requester, there is nothing wrong with the proposed criteria based on public interest, but we would like to propose additional criteria, or alternative uses of resources.
• The value of the complaints process is not in releasing any particular bit of high impact information, but in shaping the overall flow of information through the FOI system.
• The ICO is generally coming to a problem late in the process, effort spent on making the process deliver the correct conclusion earlier would help release more information that meets these criteria, without active intervention on a case by case basis.
• Prioritisation can be used to strategically attack problems with large systematic impact.
  • In general, better system wide impact may be had by using freed-up time to support use of legal enforcement powers, rather than different case work.
  • Complaints about administrative silence/stonewalling have a key impact on journalists and seem a good candidate for a fast track response. We recommend this be included as an additional factor. Appreciating recent improvements in initial response time, this is a process that could be further improved through tighter deadlines for non-legal steps, and more resources to support enforcement processes.
• Practically, the proposed criteria in combination might struggle to keep to the anticipated 10-15% of the caseload.
  • WhatDoTheyKnow would, for instance, help our users working in the public interest express that in complaints. The majority of respondents to our user survey indicate they believe their request is in some way of benefit to specific groups in society, or to their local community. Journalists would make more use of a service specifically aimed at them, etc.
• We recommend moving to a system where a certain percentage of complaints that pass a (broad) set of criteria are escalated.
  • Fast-tracking a certain percentage should be more manageable, and is adjustable based on volume, preserving the speed of the fast track, while not completely separating the two streams.
  • Sampling from a broad pool helps avoid effectively managing workflow through definitions. More loose definitions can be used, without overwhelming the system.
• As part of a broader set of criteria, we would highlight requests made in public (through, but not limited to, platforms like WhatDoTheyKnow) as being a factor in evaluating public interest - as the improved discoverability of the results means more information is available to the public in practice, as well as in theory.
• Some aspects of dealing with requests faster are in themselves useful in shifting wider behaviour. Setting short deadlines for prioritised appeals (because information should all be gathered after internal review) is a good principle that could be more widely applied to many complaints. Reducing opportunities for authorities to advance a new reason to withhold information, leads to better initial decision making.
• Fuller explanation of some of these points can be found at https://research.mysociety.org/html/prioritising-foi-complaints/#the-public-interest-and-foi

Question 2a: In particular, do you agree that prioritising cases based on who has made the request is an appropriate public interest factor? If so, are there any other groups or types of requester you think should be covered?

• No
• We disagree with prioritising based on types of requester.
  • This goes against the principle of requester-blind requests.
  • There are practical issues in validating who are journalists and relevant civil society groups.
• We do acknowledge that there may be particular reasons to bend the rule for journalists (they are especially vulnerable to delays in release of information and in practice not always treated requester-blind by authorities) - but criteria based on public interest and impact would be a better way of achieving the same goal.
• If continuing, better reframed as kinds of activity (‘journalism' rather than' journalists') for clearer definitions. This is in line with the broad US FOIA approach to defining journalists, and we would recommend something similar to that definition.
• In general, resources might be better spent on dealing with problems experienced by journalists (e.g. public interest test delays, targeting time-sensitive information, or stonewalling), either through case prioritisation or enforcement, that improve the flow of information through the whole system.
• Fuller explanation of some of these points can be found at https://research.mysociety.org/html/prioritising-foi-complaints/#avoiding-a-two-tier-system

Question 3: Do you have any comments on the service standards (or Key Performance Indicators) we should set for dealing with our FOI and EIR complaints?

• We approve of having these performance measures, and would propose several additional measures that would be useful.
• While respecting the funding conditions, the faster the ICO can respond, the less incentive there is for authorities to use the appeal cycle as a delaying tactic.
• The ICO has committed to this elsewhere, but in general making sure casework figures are regularly published is an important part in understanding the current situation beyond these headline numbers, and that there is not a sudden cliff at the six month mark.
• Additional KPIs or tracking of how many requests are triggered under different criteria would be useful in reviewing if the fast stream in practice meets expectations of around 10-15% of volume.
• Continuing the theme of the previous answer, setting a separate KPI for a shorter timespan dealing with administrative silence complaints would be a good way of reflecting the ICO's interest in the issue.
• See more in response to question 5, but setting an expected (low) KPI of complaints that will be dismissed under a new frivolous definition would be useful for understanding expected impact, and checking that against the reality.

Question 4: Do you agree that 6 weeks is sufficient time to bring a complaint to the ICO? If not, please explain why you think additional time is needed or what any exception criteria should include?

• Yes
• With indicated allowance for exceptional circumstances, this seems reasonable.

Question 5: Do you have any comments on the ICO's approach to implementing the Commissioner's statutory right to not make a decision where a complaint is vexatious or frivolous?

• Given the qualifications and grounding in tribunal decisions, we have no strong feedback on the vexatiousness definition. However, we have concerns about the frivolous definition.
• This definition is concerned with the public interest in the information under consideration. But the public interest goal of the ICO is not necessarily that every bit of information released is itself in the public interest, but that the overall system of Freedom of Information works.
• People should be able to request information that is useful for them alone. The ability for people to do this is, across the system, in the public interest.
• A modification of the definition around the "public interest in pursuing the complaint" reflects the spirit that the Commissioner has expertise in recognising complaints that, practically, are going nowhere, while recognising that the ICO's role includes defending FOI requests made by private interests.
• The ICO should commit to a KPI on a (low) volume of complaints stopping through this route, and publish these complaints so the scale of this strand can be understood.