I& #39;m kicking this again because I think that it& #39;s important to underscore: The decision to count something is a political decision. And the count of a thing is not necessarily reflective of the underlying incidence of that thing, and may be more reflective of the counting process. https://twitter.com/ProfHalloran/status/1232638494603988998">https://twitter.com/ProfHallo...
Related content: https://twitter.com/alexismadrigal/status/1234911867207745536?s=20">https://twitter.com/alexismad...
Hi, Chris, become comfortable with not knowing how many cases there are. In fact, become comfortable with not even coming close to knowing how many cases there are. Cases are really, really hard to count. https://twitter.com/chrislhayes/status/1234929424593756161?s=20">https://twitter.com/chrislhay...
Related content (indicators for system performance/proxies for incidence): https://twitter.com/NickRHart/status/1234654161897185280">https://twitter.com/NickRHart...
This thread adds an interesting dimension to the systems analysis; How fast can you get new capacity on line.

In systems (like child welfare) that run at near capacity, increasing that capacity (new beds/foster parents, new workers, etc.) is a process that has its own timeline. https://twitter.com/LizSpecht/status/1236095184170958849">https://twitter.com/LizSpecht...
And so we don’t lose the thread, these systems things aren’t just tangentially related to the case counts: these are things that are partially responsible for those counts. We can’t think of counting solely on the individual level, we must think of it contextually, ecologically.
Another reason systems thinking is important is because eventually you encounter a justice question like: ‘I wonder when we will begin asking the hard questions such as "if hospital beds become scarce, who gets in?"’

Asked and answered here: https://twitter.com/DellAnnaLuca/status/1236459870557102080">https://twitter.com/DellAnnaL...
Here’s a story about the consequences of these kinds of capacity constraints and systems effects playing out in the child welfare system. https://twitter.com/ProPublicaIL/status/1237696776708263937">https://twitter.com/ProPublic...
When resources are scarce or capacity constrained, then there is necessarily required some way to allocate those resources. This has practical and ethical consequences. https://twitter.com/GovindPersad/status/1238253335423090689">https://twitter.com/GovindPer...
Here’s a pretty clear (and simple!) example of how policies around how counts are done/reported impacts things: https://twitter.com/JustinWolfers/status/1239155894144860161">https://twitter.com/JustinWol...
Here’s an interesting illustration of how prevention strategies impact bed capacity: https://twitter.com/carolecadwalla/status/1239688968058396672">https://twitter.com/carolecad...
When we’re thinking about system capacity, it is a function (though not an aggregate sum, for various reasons) of multiple capacity constraints that exist within the system. Here’s an example: https://twitter.com/JoeKristol/status/1240244523164471296">https://twitter.com/JoeKristo...
And I just realized that I’ve hopelessly muddled all of these different systems threads... Here’s another one that should be joined to this: https://twitter.com/ProfHalloran/status/1235298459026104326">https://twitter.com/ProfHallo...
And here& #39;s an interesting thread on ventilator capacity: https://twitter.com/CitizenCohn/status/1240262582780473347?s=20">https://twitter.com/CitizenCo...
This is an article on the difficulty of achieving accurate case counts. A must read. https://twitter.com/BuzzFeedNews/status/1242994053794365440">https://twitter.com/BuzzFeedN...
And an intensely political part of the problem with obtaining case counts that reflect the scale of the problem: https://twitter.com/MarcACaputo/status/1243234188045795328?s=20">https://twitter.com/MarcACapu...
Another really interesting article on case counts/undercounts. https://twitter.com/GlennKesslerWP/status/1243528639330820096?s=20">https://twitter.com/GlennKess...
This is a really good (not at all boring) thread on the ways that we can think about count data that& #39;s related to COVID-19. https://twitter.com/politicalmath/status/1243950119684206598?s=20">https://twitter.com/political...
Undercounting both cases and deaths was inevitable, and I’m glad that the problem of counting is receiving attention in the Times.

Remember counts are political. Remember counts say as much about the process of counting as they say about the underlying phenomenon. https://twitter.com/sarahkliff/status/1246863725438664709">https://twitter.com/sarahklif...
Another piece on indicators and undercounting: https://twitter.com/mckaycoppins/status/1247480287807836161">https://twitter.com/mckaycopp...
Yeah, one of the biggest unknowns right now is the prevalence of coronavirus in America. And it’s an unknown because a lack of wide scale testing prevents us from even being in the neighborhood with our estimates about number of cases. https://twitter.com/theplumlinegs/status/1248937990786822145">https://twitter.com/theplumli...
This thread is a look at how deviations from expected mortality trends can help us evaluate death counts in a historical context. https://twitter.com/sangerkatz/status/1248685404775239680">https://twitter.com/sangerkat...
An interesting read on how hospitals can increase capacity: https://twitter.com/RANDCorporation/status/1249870991225913344">https://twitter.com/RANDCorpo...
Remember when we were talking about how the decisions whether to and how to count were political? https://twitter.com/ImmCivilRights/status/1250806395383042049">https://twitter.com/ImmCivilR...
It me.

Sigh. https://twitter.com/Noahpinion/status/1251343070513131521">https://twitter.com/Noahpinio...
Thread:

https://twitter.com/andrew_benesh/status/1255171160511254528?s=20">https://twitter.com/andrew_be...
Again, I think the undercounting of COVID deaths has a whole lot of explanations (most of which don’t include the phrase cover up) but the WaPo and NYT stories about excess deaths raise lots of questions about the range of COVID mortality. https://twitter.com/No_Little_Plans/status/1255677399150268417">https://twitter.com/No_Little...
Here’s the WaPo article: https://twitter.com/WeinbergerDan/status/1254802347894558720">https://twitter.com/Weinberge...
Here’s the NY Times article: https://twitter.com/nytimes/status/1255502123644661768">https://twitter.com/nytimes/s...
Again, counting isn’t simple. Even counting deaths—which, in a way, seems obvious—isn’t simple.

To understand what a count means we need to understand what is counted and why. https://twitter.com/JimPethokoukis/status/1256022722745901056">https://twitter.com/JimPethok...
Tangential to COVID (though close to my area of study), but important and illustrative of how the process of counting impacts what we define as the count. https://twitter.com/benconomics/status/1256022077917720576">https://twitter.com/benconomi...
And a gentle reminder that a critical view of count indicators (which we should all adopt) doesn’t mean that indicators themselves are worthless: https://twitter.com/katzonearth/status/1255962301783515136">https://twitter.com/katzonear...
This is not strictly about counting, but about modeling counts (honestly, though, one of the reasons talking about counts is important is important is because it informs the way that we interpret models). https://twitter.com/dwallacewells/status/1256579601498869761">https://twitter.com/dwallacew...
As important a question as “what is the count” is that of “why is there a count.” https://twitter.com/Ethnography911/status/1257466901611130880">https://twitter.com/Ethnograp...
I don’t want this thread to drift too far into modeling (though count quality is an essential component of modeling), but this is an important read on how data is consumed.

I include it here, because that’s really what contextualizing counts is all about. https://twitter.com/claudiakincaid/status/1258730173924589570">https://twitter.com/claudiaki...
When I started this thread about counts being political, I wanted to call attention to the subtle ways that information gets categorized and organized, along with the difficulties if counting. But sometimes count politics are more foregrounded: https://twitter.com/ehananoki/status/1260549784101847042?s=20">https://twitter.com/ehananoki...
You can follow @ProfHalloran.
Tip: mention @twtextapp on a Twitter thread with the keyword “unroll” to get a link to it.

Latest Threads Unrolled: