Lectures

Predicting Across Time and Space

This article is part of the series:

Every day for the month of March 2021, I watched the numbers of Covid-19 cases in Kenya climb and climb. Family and friends sent me messages that the cases were rising. Friends and friends of friends started falling ill. The graph of Covid-19 cases in Kenya at the visual dashboard in the Covid-19 Data Repository by the Center for Systems Science and Engineering (CSSE) at Johns Hopkins University showed a rising line that begins to crest in April 2021. Surrounded by the language of waves and surges from the Health Cabinet Secretary in Kenya, Mutahi Kagwe, newspapers, as well as local and international news about South Africa, Tanzania, the UK, the United States, Brazil or Italy, people around me talked of a third wave in Kenya. And the government seemed to be doing nothing. Nevertheless, a friend of mine, who had tested positive for Covid-19 in Nairobi, raised the provocative question, “but have we really had our first wave?”

I find this question intriguing. What is a wave? The visual dashboard in the Covid-19 Data Repository has three distinct bumps in the graph indicating increased Covid-19 cases in Kenya: July through August 2020; October into December 2020; and then March through April 2021. Each bump is a little larger than the last and each with a rather rapid dip after the government restated or increased lockdown and other Covid-19 mitigation measures, such as restricting large group gatherings (political rallies, specifically), extending curfew hours, and mandating mask-wearing. In addition, according to the World Health Organization (WHO), by April 27th there were 156,981 confirmed cases in Kenya and 2,643 confirmed deaths. However, to call this a wave, one would have to arrange these case numbers in relation to case numbers of the past, and establish some connections between these numbers and the cases yet to come. It is this past and future together that form an entity that could be defined as a wave.

In this piece, I ask how we might give significance to the relationships between past and future numbers, and between local and distant cases; and to how these numbers and cases have been transformed into predictions about Kenya and Africa more generally. I argue that predictions, if made at all, cannot be abstracted, narrow, or universal. Rather, they must be particular and parochial, yet expansive enough to fit the needs of specific local dynamics. I also suggest that we look at how, what, and why information is compiled into predictions and given significance. As such, I interrogate the idea of being able to know a wave or a surge. I also question how that knowledge is constituted and then used for policy in light of failed past predictions.

Back in February 2020, at the beginning of the awareness about Covid-19, the heads of the African Center for Disease Control and Prevention, John Nkengasong and Wessam Mankoula, singled out Africa as a space of greatest concern in facing a pandemic, citing weak health care systems, inadequate surveillance and laboratory capacity, scarcity of human resources, and limited finances. Kenya was assessed to have moderate risk with variable capacity to respond and high vulnerability from a virus that at that time was predicted (and therefore modeled) to arrive in Africa from China. By October 2020, however, Kenya appeared to have different Covid-19 dynamics than what had been anticipated. Clinician Jon Fielder, commenting on the dynamics in Kenya, specifically, and Africa more broadly, argued that dire consequences were originally predicted because Kenya’s population is medically vulnerable due to conditions like HIV, tuberculosis, heart disease, and hypertension and to the fragility and limited capacity of the medical systems. These underlying conditions have not changed. Nevertheless, the predicted total devastation has yet to be seen.

Many theories have been put forward as to why this is so, including that African populations were “used to” pathogens and thus more able to fight them. This of course contradicts the previous claim that Kenyans are more medically vulnerable due to high rates of other diseases. There have been many other explanations proposed for the differences between Africa and the rest of the world regarding Covid-19 consequences, including arguments about genetics, demographics, pathogens, inoculations, warm weather, and the idea that much of the population was already exposed to or had already contracted Covid-19. Similar arguments were used as recently as November 2020 and February 2021 to explain why India was less affected than Europe or the Americas. At least until March, anyway. What a difference that a month makes. With skyrocketing cases in India, analysts of India have had to relook at these assumptions about immunity.

In early March 2021, I travelled to the United States and, while there, saw a doctor. At the doctor’s office, I was required to fill out a form that asked: “In the last two weeks have you been in a place where Covid-19 was surging?” I hesitated for a minute. Were we surging in Kenya? By March 4th, there was a noticeable uptick in the number of cases, but was it there when I left on March 1st? Was this a surge? Was it there in the past two weeks? I couldn’t really say for sure. The next few weeks would see an alarming rise in cases followed by a pronouncement by the Kenyan state of increased mitigation measures on March 26th and then a significant drop in numbers of cases.

Yet, in early March, was it a surge? If we take into account the cases over the last year, have we yet had the first wave? How would we know? The problem with the idea of a ‘surge’ is that it is a forward prediction and nothing about Covid-19 in Kenya (or Africa) has been predictable. How do you make governmental decisions in unpredictable times? Of course, these times are unpredictable everywhere. Yet, similar to the weather, one knows that although it is not very predictable, in some countries weather forecasts allow you to dress without getting wet. Those forecasts are pretty accurate over a small geography. Not so in Kenya, where they are generally wrong on a daily basis and even for small geographical areas. In Kenya, weather forecasts are generally right only in the most generalized terms. For instance, the weather forecasts in Kenya in the 1990s used to say “scattered showers and thunderstorms in the afternoons east and west of the Rift Valley” almost every day of the year. Or, to put it another way, ‘somewhere east and west of the Rift Valley it will rain at some point in the afternoon.’ This may or may not be where you are, even if you are east and west of the Rift Valley, but rain will happen somewhere. Basically, this is as useful as the saying “it is Happy Hour somewhere right now.” It is both true and unhelpful for your particular place. But, I argue, this does suggest something about predictability, specificity, spatial dynamics, and making one statement that is meant to cover broad areas of time and space.

If we say these are “unprecedented times,” this is a past projection that we can all agree on. But a surge—or the recognition that these are not waves and the big one is coming—requires that forward pattern recognition that what we see now, and what we see elsewhere, can tell us something about what we will see here, later. It is a problem of both time and space. This is also part of the problem of knowing whether or not we have had a first wave. If looking at a data visualization graph, it looks like distinct waves of Covid-19 cases. If looking at elsewhere in the world, say the US or Mexico or France, it doesn’t look similar at all. Are the specifics of what is happening in Kenya related to phenomenon on a global scale? Is it even an issue of scale?

In the case of Covid-19, this predictability and relationality across time and space has not really worked in regard to Kenya or most of Africa. When the rest of the world was so hard hit in the first wave, we were sure, given the limited resources, the lacking health services, the small economies, and the supposedly inept governments, that this would mean total devastation for Africa. This is in spite of differences in speed of lockdown, curfew, mask wearing, and governmental regulation; in the continent-wide coordinated response; in the uptake of sanitation measures; in the demographics of age, density, obesity, and in the urbanization of population.

When this anticipated devastation did not come to pass in Kenya, we thought that with lockdown partially lifted and the exodus from cities, national spread would happen. Sure enough, a second wave hit. But its slow take off and fast turn down was a surprise (in spite of speed of government response, remaining curfew, mask wearing, and the demographics of age, density, obesity, and urbanization of population). Then came the threat from the south, midway through March 2021. A new variant. Sick people crossing borders. Members of East African governments falling ill. A surge? It definitely looks like a surge.

On March 26th 2021, I was incensed that it took Uhuru Kenyatta’s government so long to finally see what I understood as a surge–three weeks too late. In his presidential speech on March 26th, Kenyatta admitted that whatever they thought on March 12th, they no longer believed two weeks later. Masks were demanded. Curfew extended. Schools closed. Public gatherings restricted. Political rallies banned. Lockdown again. Nairobi back as a containment zone. Our positivity rate was at 22%. I imagined we would end up like South Africa, the UK, or Brazil, with their uncontrollable variants. But, this didn’t happen. Rather, after increased Covid-19 mitigation rules put in place by the government, the curve quickly turned down. These mitigation measures were relaxed on May 1st as cases continued to decline.

There was substantial pressure on the government to relax mitigation measures and to own up to their part in the spread of Covid-19. In particular, political rallies connected to the Building Bridges Initiative (BBI)–a proposal to change the National Constitution by those in power–were seen as contributing to a surge in numbers. As one social media post shared on WhatsApp by “Depressed Kenyan” put it: “I work as a waitress at Alfajiri grill, Kilimani. Prior to surge in numbers there were political gathering all over in the name of pro & anti BBI. Unfortunately, me and other low-class Kenyans are paying for this carelessness by losing our jobs #UnlockOurCountry.” Depressed Kenyan and I saw the same problem of politics increasing Covid-19 cases. But not the same solution. Looking at their jobs, they wanted to unlock the country, and looking at rising numbers, I wanted to re-lockdown. For, again, a problem is a look to the past and a solution needs forward projection, but we were solving different issues caused by continuing political rallies.

In April 2021, even the Kenyan Interfaith Council decided that cases were surging and canceled celebrations of Easter, Ramadan, and Hindu feasts. So yes, there must be a surge, right? After all, in the West religious institutions seem incapable of responding to the pandemic in any way that takes the life of their congregants into account. But then, is it only life that counts? What about quality of life or sanity or sanctuary that in theory religious institutions provide? Who is to decide?

So, are we surging? With new ultra-contagious, more deadly variants? Well, yes and yes, but…there seems to be a downward trend of the curve now a month later. So fast? Is this wishful thinking or an actual turn? And as the weeks pass, the turn continues to trend downward. Why did we not follow a trajectory like India, which is out of control in the opposite direction? How much difference a few days make. How do we make sense of something that doesn’t seem to behave the same in different places? How much is this a result of geography? Demographics? Sociality? Government response? Religious response? Does it matter if we and our neighbors count cases or not? What are the protective networks and the deadly ones, and to whom? And, finally, what is the cost of this African unpredictability in the virus for the economies of African countries and for the healthcare systems of the world? How do we plan?

You see, poverty also kills. Paul Gubbins of Financial Sector Deepening (FSD) Kenya points out that two million people in Kenya were reduced to poverty in the process of last year. Two million! Our hospitals were not yet overwhelmed but our economy was hit hard and the effects disproportionally felt in terms of poverty and education. Looking back, should we have done things differently from the beginning? How should we have decided what information to give significance to when building local forecasts? Should we have not believed that what has happened elsewhere will happen here? Or was the difference because we took faster, earlier measures (in virus time) as we had a bit of warning being off the beaten track of world commercial networks? Did this buy us a year? Did our policy response have a problem of believing narratives about Africa or our governments or our untouchability that stopped us from doing a very different—but equally effective—virus prevention with less economic consequences? Again, only in hindsight does any of this become obvious. What is needed to make predictions? Are we in a surge? Is this a real wave? How would we know?


Bettina Ng’weno, an associate professor in African American and African Studies at the University of California, Davis, is from Nairobi, Kenya. Trained as an anthropologist at Johns Hopkins University, she studies space, citizenship, cities, and states in Latin America, Africa, and the Indian Ocean region. Working from personal, familial, ethnographic, and archival history and experience, her current book and film project brings to life a Nairobi centered on the railways, the dreams and aspirations of long-term residents, and the complicated spatial and temporal dynamics of the city. Recent publications include Reimagining Indian Ocean Worlds (Routledge) and Developing Global Leaders: Insights from African Case Studies (Palgrave).



Leave a Reply

Your email address will not be published. Required fields are marked *