My last post continued a series of posts about my suggest to impose a rigorous system of activity tracking on all employees and contractors of the government.
Maintaining voluntary democratic consent of bureaucratic government provides the justification for activity tracking. This is not a minor issue. The recent Ferguson riots and protests demonstrate how quickly and violently the population can reject government even when the government makes no violations of any written laws or duties. Among the suggestions to avoid repeating the Ferguson problem is the suggestion to equip police with body and vehicle cameras to record the future confrontations. We missed the opportunity to have cameras in this case because we did not pro-actively adopt these technologies. With the lessons of Ferguson, we are approaching a consensus that they are a necessity.
Such cameras are similar to activity tracking in that both give the appearance of micromanagement. Other than cost, training, and burden (more equipment on the body), a major objection to their use is that the cameras are very intrusive. I suspect the latter is the primary reason for the reluctance of the police force to accept this technology.
Although the Ferguson episode teaches us how the conditions can suffer without these recordings, I suspect there is still reluctance because the cameras would also capture many instances that could be used against officers in non-controversial cases. Controversial confrontations represent a small portion of their duty lives so that the vast majority of the recordings will be of routine activities. It is possible that the recordings from body and vehicle cameras can create more controversies than they solve.
I have not studied the actual protocols for handling recordings from body and vehicle cameras. Maybe there will be a policy to immediately seal any recordings and not release them unless there is a controversy.
My point here is that workers object to intrusive recordings of their daily activities for concerns of how their routine activities (not associated with a controversy) could be used against them or could be misinterpreted. Generalized, this is a complaint of micromanagement. People generally prefer to work autonomously once they have qualified for a job.
Police in particular require extensive training and certification to begin their duties. Historically, they have been given reasonable autonomy to decide for themselves the best actions to take. They may not like the recording of their activities., but the public is demanding it. The public is aware there are practical technologies that can help clarify what happened in cases like the one in Ferguson. With that awareness, the public will demand its use.
My proposal for activity tracking of bureaucratic work is to record metadata about how each worker invests his duty hours on actual work products. My proposal may appear similarly intrusive as police body cameras. However this proposal uses coding of categories activities to distinguish the nature of the effort involved. These codes would not reveal the details of the content of that work and certainly not to the level of detail as video recording of duties.
The goal of the coding is to be precise enough to enable analytics to allow the public to learn that the right amount of effort was invested by the right types of labor at the right time to justify some government product such as a new regulation. The coding would be abstract enough so that this data can become immediately available to the public for ad hoc queries without the need for submitting FOIA requests that invoke manual efforts for approvals.
This proposal also appears similar to time-keeping. Current time-keeping involves job numbers that cover a wide range of activities over a long period of time so that often an individual can assign all of his time to just one number. The new proposal does not replace time-keeping. Time keeping will continue as it is. Instead, this proposal is to add an independent requirement to track all distinct job-related activities that occur while the worker is on duty. For internal quality control, there may be a check that the activity tracking is consistent with the time-keeping, but the intention is to keep both systems independent of each other. They serve separate functions.
For democratic visibility into the work done inside government, we are interested in the record of the actual work efforts invested. As I mentioned in earlier posts, these activity codes would capture the nature of work performed (for example, separate codes distinguishing conversations from phone calls and phone calls from e-mails), the justification for performing the work (for example, to identify the specific departmental-level task), and whether the task is an initial request or a particular reason for following up on an earlier task (for example, a revision). In my last post, I noted that it is impossible to itemize all of these codes in advance so we need a flexible approach that allows each individual to propose new codes and allows for a process to collect these proposals into unified codes for similar activities.
Distinguishing time-keeping from activity-tracking allows us to propose an independent technology specifically for tracking activity. The more detailed recording of work efforts desired for activity tracking may involve recording many (perhaps dozens) of activity codes every hour.
To make activity-tracking possible, we need an approach that is different from time-keeping that usually involves just a couple codes per day but requires manual signature to attest to the accuracy under penalty of discipline if it is incorrect. To the largest extent possible, the activity-tracking needs to be automated so that the activities can be recorded without burdening the worker at all. Where there is a need for manual coding, I imagine leveraging mobile communications technology (portable electronic devices like smart phones) to allow a person to quickly select from a personalized list of common codes. For this discussion, I will refer to these as PATs (personal activity trackers) but I’m referring to something like a smart-phone or a pad computer.
Automated approaches may include computer workstations logging what applications are active while a person is logged in.
Most of activity tracking may only be partially automated. For example, a worker may make a phone call and there can be an automatic record of the amount of time spent on the phone call. However, the activity-tracking code to describe the nature of the call can not be automated. The technological solution may be to send a text-alert to the employee’s PAT to ask the employee to code the just-finished phone call. Phone calls alone may have many qualifying codes. The codes may include personal calls or friendly networking calls not directly related to work. Even work-related phone calls can have variations such as talking directly with one other person, participating in a conference call, participating in a listen-only meeting, or spending time on hold such as waiting in a help-desk support call queue. It is easy to imagine calls that do not match existing codes and this is an example for the need to allow the worker to introduce a new distinct code for this one call.
Another semi-automated approach may be for the PAT to send a text message after some period of time to verify that the person is still working on a task. For example, a person may start writing an email message to respond to a particular question. If historic data predicts that such e-mails should take about 5 minutes of time, the PAT would ask for confirmation that the worker is continuing that same activity after that time has elapsed. The PAT includes an automated polling feature to request for what the person is working on right now if he has not recorded any activity for a period of time.
I imagine there will be a way to turn off this PAT by declaring some period as an off-duty period. For example, a lunch break is off duty. Putting the PAT in off-duty mode will avoid the need to record anything during this period. Having the PAT in on-duty mode will obligate the employee to record what he is working on at that moment of time. There are codes for personal tasks, if those are needed.
As mentioned, there may be internal quality control checks to verify that the PAT’s record of on-duty time matches the recorded hours in time-keeping, and that the time spent for personal reasons is acceptable. For my proposal, I distinguish the activity tracking from time-keeping and performance reviews. The purpose of activity tracking is not to measure individual performance. Instead activity tracking is to provide the public data about the diligence within the department to meet its obligations in performing its missions.
Although activity tracking has a distinctly different goal from time-keeping, the two appear to be the same thing to the employee. Both activity tracking and time-keeping asks the employee to account for his time. Even with automated and semi-automated tracking approaches, the employee will see this as a much more intrusive form of time-keeping. From an employee’s perspective, activity tracking is an extreme example of micro-management.
Even in current practice where the focus is entirely on internal time-keeping, there are many complaints of micro-managing of staff. The supervisor may appear frequently to verify that the person is doing some task, or he may inquire if a person is using time inefficiently. This type of careful watch of time-utilization comes from a perspective of diligence in assuring that duty time is not being wasted. Wasted duty time is an issue within government and attempts to minimize this waste involves some form of micromanagement. The boss will arrive periodically to check the employee is still working on the assigned task, not just for confirmation that the task will be completed on time, but also that the person is not free to do other tasks.
Such micromanaging causes complaints and degrade morale. My proposed activity tracking approach appears to be an extreme form of micromanagement. Activity tracking involves frequent automated polling on the current task or automated demands to describe the task with codes the identifies what is involved and why it is necessary.
This proposal of activity tracking government work is very intrusive. Many (if not most) workers will object to this and protest to the requirement to do this. Even if we can fully automated activity tracking such as my previous post’s example of the sleep-tracking bracelet, the awareness of this level of activity tracking will be unacceptable to most workers. Although the purpose of activity tracking is to foster democratic consent of government by providing public visibility of the nature of work performed in government agencies, individual workers will perceive this tracking as indistinguishable from micromanagement. This level of micromanagement will be intolerable to most workers.
I recognize this but I argue that this is a necessary cost for government bureaucracies that have earned independence from day-to-day control by democratically-represented congress. To maintain democratic consent to be governed by bureaucratic actions, we need to have some method of democratic participation in the bureaucratic process. I am proposing a feasible approach to restore that democratic participation by providing access to meta-data describing the work performed by government staff.
In particular, the public should be able to learn the entirety of the actual bureaucratic work effort went into some regulation or ruling that affects their lives or that causes their concerns. In my last post, I pointed out to a recent EPA proposal to set tougher standards for ground-level ozone levels. This ruling is controversial because it follows an earlier standard for the same pollutant.
We recognize that the agency has this authority and this authority is independent of the currently seated congress. Once EPA finalize the rules, we will have to follow them (paying the costs of meeting a higher burden) with no option to complain.
This is an example of regulation that is outside of control of democratic processes. We set up the EPA to be staffed with specialists to consider complex topics. We may agree that the topics of the control of air pollution is beyond the public’s ability to solve through democratic legislation. We will also recognize the added burden of the regulation. We want to know that this burden is justified and is not capricious. My proposal of making publicly available detailed activity tracking data inside the bureaucracy will provide the population better information to give them confidence that the agency invested appropriate amounts of internal deliberations prior to issuing the new ruling.
Activity tracking restores some level of democratic participation in an increasingly bureaucratic government. Activity tracking happens to impose micromanagement burden on the bureaucracy’s employees and its contractors.
In order to be successful, a program of activity tracking of government employees would need their good-faith cooperation to capture useful data. This will require overcoming the natural objections to the micromanagement of their daily activities. I doubt that voluntary compliance will ever be fully successful because people naturally resent being watched too closely even when it is justified to maintain social order by providing democratic insight into the independent bureaucratic processes that are not otherwise accountable to democratic participation.
We may be able to address some of the complaints about micromanagement.
One complaint is that real work effort rarely matches the options available to describe current activity. In last post, I gave example of ICD-10 having one code for Ebola (EVD) but popular news accounts described 5 different scenarios attributed to EVD where only one of these scenarios closely matched the official definition. Being forced to use an inappropriate code for a task is very frustrating.
I recall an example from a long time ago is when I had a task of writing a document. I encountered a computer problem that prevented the software from saving the document. Addressing this computer problem was essential to completing the assignment, but this effort had nothing to with the concept of writing a document. If the boss asked “are you still working on the document” the correct answer “no, I’m trying to fix the computer” would not have been a valid option. The right answer is “yes, I’m working on the document”. The additional clause of “but first I need to fix the computer” would be ignored even if I mentioned it. This is frustrating.
Another complaint is that the description of the task does not match the actual task at all. An example may be to have the task to prepare for a meeting but the official task description discusses preparation of briefing slides while in this particular case the intention is to talk without any slides.
Another complaint is that the task is not justified. An example may be take a mandatory training course on ethics of entertaining clients when a particular employee will never have direct interaction with clients. Here the frustration comes from the inability to describe that task as irrelevant even if it were mandatory.
I already mentioned one step to alleviate the appearance of micromanagement by separating activity tracking from existing time-keeping and performance reviews. We need to emphasize that detailed activity tracking serves a separate purpose from evaluating individual performance. Activity tracking serves the needs for modern democratic participation in government by providing readily accessible evidence that the bureaucracies are performing due diligence in their actions.
Another step to alleviate this frustation is discussed in my last post where I described creating a coding system that permits the workers to create new codes to more accurately reflect the actual task they are encountering instead of being forced to use an irrelevant code. That design used GUIDs as keys to database records that will provide multiple dimensions about a task. This will permit a person to create a new code (with a unique key) that copies most of the dimensions of a previous code but replaces the content of one or more dimensions to be more appropriate to the current task.
Personally I would find this type of flexible to be satisfactory. I was very eager to describe in detail of my activities while on the job as long as I was able to describe the actual tasks accurately. I demonstrated this by writing lengthy and detailed monthly activity reports that attempted to exhaustively list the activities I performed. Even though these were presented as prose, I recognized that it would be very helpful if I could code the activities in such a way to more easily queried in a database. Even for just this one worker (myself), it would be handy to query over the previous ten years how many times I did a certain type of task and how long it took over time. I did not find this process frustrating or too intrusive because I had complete control of the descriptions of my activities. All of my activities were written out in prose.
I was disappointed in that my example did not inspire others to write similarly detailed accounting of the work they actually performed. Although I did not rigorously explore the reasons for their reluctance, I did learn from some discussions that people resent having to report at such detailed levels. They are only barely tolerant of the annual review process that asks only for a brief description of their activities and accomplishments over the prior year.
Realistically, in order to get the level of activity tracking required for my proposal, the law must obligate this tracking through automated logging (where feasible) and semi-automated polling to request worker confirmation of the type of task he is currently performing. Such legally obligated requirements have precedents. These are similar to the reporting requirements that government agencies impose through regulations on commercial industries. The government itself can operate under similar reporting obligations it imposes on industry.