Gideon's blog

For a recent work project I have been using openrouteservice and the Vehicle Routing Open-source Optimization Machine (vroom) to conduct a route analysis for a client interested in evaluating new warehouse locations for shipping. Configuring this software gave me a huge amount of trouble - I was finally able to get things working using the info below, which I also posted at https://github.com/VROOM-Project/vroom-docker/issues/27.

As an absolute novice at Docker I suffered for many hours trying to set up Vroom + openrouteservice and finally got my installation working.

Here are the main things I needed to fix to get this to work:

ors and vroom must be on the same Docker network to talk to each other. You can do this in two ways: either start both containers separately, creating a new docker network via the docker network create command, and then connect both containers to the network using the docker network connect commend. even better, add vroom to the ORS docker-compose file. I added the following lines to my docker-compose file:

# ----------------- Vroom application configuration ------------------- #
  vroom:
    image: ghcr.io/vroom-project/vroom-docker:v1.14.0
    ports:
      - "3000:3000"
    volumes:
      - ./vroom-conf:/conf
    environment:
      VROOM_ROUTER: ors

The default port in the Vroom config file did not work for me. I had to change the port from 8080 to 8082. I troubleshooted the port by opening up a terminal within the Docker vroom container and running curl requests at the ors-app container until I found a port that connected - I'm sure there is a better way. I also had to change the name of host to "ors-app" to match the name of my ORS docker container. In my Vroom config file I ended up with the following lines:

routingServers:
  ors:
    driving-car:
      # host: '0.0.0.0/ors/v2'
      # port: '8080'
      host: 'ors-app/ors/v2'
      port: '8082'

This gave me a giant amount of trouble but I now have Vroom and ORS running locally. For other novices, I highly recommend following the Docker tutorials and paying special attention to the "multi-container apps" tutorial. I really wanted to avoid having to learn Docker for this but it ended up helping me understand the issues I was facing.

Originally written as memo project for Quant II at Columbia SIPA, as response to prompt to evaluate MFI research for a hypothetical World Bank project.

As the World Bank considers expanding its microfinance operations, we must first evaluate the evidence on whether microfinance actually has positive impacts on its recipients’ material well-being.

Three recent studies shed light on microfinance’s impacts as measured by wealth, income, and expenditures: Banerjee and Duflo’s experimental 2010 research in which 104 slums in Hyderabad, India, were randomly selected for the opening of a microfinance institution (MFI); and Kondo (2008) and Montgomery’s (2015) two quasi-experimental studies that attempt to remove bias from observational data using a difference-in-differences model.

Findings from the three studies are mixed. The strongest and most consistent impacts are on household expenditures related to starting or improving businesses, and related entrepreneurial activity. Banerjee and Duflo found that in MFI treatment areas, there was statistically significant evidence of a shift in expenditure composition, with increased spending on durables and reduced spending on “temptation goods” and festivals. Montgomery found that MFI “yields the most impact for urban households running microenterprises and for very poor borrowers engaged in agriculture.” And Kondo found a “very significant positive” impact on household business enterprises. All of these findings suggest that, for households with a propensity to invest in a business, microcredit may spark increased business expenditures with the possibility of leading to longer-term material improvements for both households and communities.

Results concerning broader impacts on income, wealth or expenditures are less consistent. Banerjee and Duflo found no significant impact of MFI on total expenditure amounts. Kondo found “a mildly significant positive impact on per capita income, per capita total expenditure and per capita food expenditure,” yet this impact was regressive: benefits were found exclusively among the richest quartile of households, and in fact poorer households saw negative effects. Montgomery’s Pakistan results - despite a rosy conclusion that MFI participation “has positive impacts on both economic and social indicators of welfare” - is also fairly weak. The only statistically significant impact on expenditures was that the study’s “core poor” group, or households in the bottom quintile of the population, increased educational expenditures - yet this was in comparison to average MFI participants who were found to have a statistically significant reduction in educational expenditures compared to non-participants!

Assessing these impacts is complicated by various limitations and flaws in the studies reviewed. Banerjee and Duflo look at a relatively short time period of 15-18 months from the introduction of MFIs to the final survey. This short study length would fail to reveal any long-term expansion of consumption, as theoretically microfinance recipients may reduce short-term expenditures in order to save for longer-term plans; and it would not reveal long-term negative impacts such as indebtedness arising from microfinance loans.

Kondo’s study has a potentially serious flaw. The sampling of non-participating households in both “treatment” and “control” villages was not drawn from a random pool of villagers, but rather from lists of households “identified by MFI field personnel, center or [village] leaders.” This raises the possibility that samples were drawn from a population intentionally skewed by MFI staff in order to overstate the effects of microfinance.

The difference-in-differences model used by Kondo and Montgomery also has potential design issues:

• The studies’ use of “treatment” and “control” villages assumes similar characteristics in MFI participants after controlling for village fixed effects – but people selected in the earlier stages of an MFI program may differ from later selectees in ways that are not controlled by village fixed effects. If MFIs relaxed their selection process, or people selected earlier are for some other reason more responsive to microcredit, this could lead to overstated findings.
• Being selected for an MFI program may affect people’s behavior in ways that overstate impacts. If someone selected (but not yet receiving) microfinance cuts back on spending to prepare for investing in a business, that could narrow differences in the control group and lead to overstated expenditure impacts.

All three studies must also be evaluated in terms of local contexts, and great care should be taken before applying findings to new settings. For example, Banerjee and Duflo’s study is from Hyderabad, the capital of the Indian state where microfinance has expanded the fastest. Not only are the people there likely familiar with the microcredit model, they are also entrepreneurial: As the study puts it, “31% of households ran at least one small business at the baseline, compared to an OECD-country average of 12%.” And the Kondo and Montgomery studies both depend on data from villages where MFI programs either already existed or were planning to move, meaning that impacts may not hold for villages not already selected for an MFI program.

In conclusion, these studies provide some helpful evidence for considering microfinance expansion. The strongest evidence from these studies is that microfinance can be an important tool for improving material well-being in places where people are predisposed to make business investments and where microfinance has an existing foothold. The fact that this finding emerges both from a well-designed short-term experimental study, as well as two longer-view quasi-experimental studies, is quite promising.

On the other hand, we should be skeptical that microfinance, by itself, can raise basic living standards. The three studies have weak and mixed results on microfinance’s ability to raise baseline consumption or expenditures, and there is some evidence of microfinance causing harm. Additionally, all three studies take place in contexts where microfinance already exists, and all depend on people self-selecting their membership in loan programs. There is no evidence presented here that creating a microfinance program in a randomly selected poor region, or pushing people to join an MFI, would have positive impacts.

Microfinance may be able to play a role in helping some households in specific circumstances, but this research does not suggest that microfinance can necessarily expand basic material wealth without strongly considering population and implementation.

Originally written for Politics of Policymaking Fall 2021 at Columbia SIPA

Policy tool: Randomized controlled trial - health effects of electric induction stoves

Recent research has revealed health problems associated with the use of gas stoves, which emit nitrogen oxides, carbon monoxide and formaldehyde, all of which can have negative impacts including respiratory issues. (Seals, 2020) An observational study from Australia found that approximately 12% of all cases of childhood asthma are attributable to gas stoves. (Knibbs, 2018)

In response, various organizations have proposed policies to drive the adoption of electric, induction stoves, which produce less pollution. For example, the Rocky Mountain Institute (RMI) recommends that policymakers “[p]rovide financial incentives, such as tax credits or rebates, that will enable low-income homeowners to eliminate gas stove pollution, including adding plug-in induction stovetops or switching from gas to electric stoves” and that such policies “[p]rioritize homes with children and other at-risk populations.” (Seals, 2020)

While there is compelling evidence that gas stoves pose health risks, no studies to date have assessed health outcomes from replacing gas with modern electric induction stoves.[1] Observational studies such as the Australian study cited above can estimate pollution impacts but cannot account for all lurking or confounding variables.

A randomized controlled trial (RCT) could directly study health effects of replacing gas with induction stoves, and provide support for subsidy policies while providing insight for effective targeting of vulnerable populations.

Key considerations in designing an RCT include finding a relevant population and collecting appropriate data.

The RCT population should be selected in line with RMI’s proposal for targeting rebates at vulnerable populations. Therefore, this RCT should study children with asthma living in households with a gas stove.

If cost is no object, administrators could attempt a national study. A more economical option would be a regional study of participants identified through collaboration with New York-Presbyterian Morgan Stanley Children's Hospital, a premier children’s hospital and one conveniently affiliated with Columbia University. Patient selection could follow a protocol similar to that used in Fiks et al., using electronic health records to identify children where asthma is a primary or current health concern. (Fiks, 2015) After working with the hospital and families to find willing participants they would be divided into control and treatment groups.

Data collection would consist of a preliminary survey, followed by health and air quality data collection before and after the health intervention. The preliminary survey would cover indicators related to health, cooking practices (such as frequency of stove use) and building attributes (such as air quality and flow, and kitchen features including exhaust fans).

Following the initial survey, both groups would be observed for a period of time - perhaps six months - before the treatment group would be provided with induction stoves. This initial period would establish a health and air quality baseline. Following the installation of induction stoves for the treatment group, both groups would be followed for another similar length of time. Key information to collect from the ongoing survey would include incidence of asthma attacks and hospital visits, symptom severity, and tests of in-home air quality. Precise study length would be determined by respiratory health experts, although minimally the study must last long enough for families to become comfortable with new cooking equipment and for health effects to be observed.

A viable study must take into account ethical and economic considerations. In working with children, administrators must obtain informed consent from both children and parents, and identify participants without violating privacy. Administrators should also be aware and consider the costs of withholding treatment from the control group. Given the widespread use of gas stoves, however, control group participants will likely not face acute risk.

The primary economic considerations are equipment and study length. Lower-end induction stoves cost around $1,000.[2] For a study with 100 participants, including installation costs, this cost would be non-trivial but not necessarily prohibitive. Study length and staff time needs could be relatively low, as asthma symptoms react relatively quickly to changes in conditions - weeks or months, as opposed to years - and the study could take place in just one region. (Jaakkola, 2019)

Following completion of the study, there are two key policy questions to consider:

1. Are induction stoves associated with changes in health outcomes for vulnerable children, and if so what are they?

2. Are there factors that affect these health impacts, such as housing characteristics, age, or severity of pre-existing conditions?

Both questions are of relevance to policymakers. The first can help gauge whether offering rebates for induction stoves (or other policies to encourage adoption) are an effective tool for protecting public health. The second can help policymakers understand whether certain populations - such as children of certain ages or residents of specific housing types - benefit the most, in order to craft an effective targeted policy.

References

Adane, M. (2021). Biomass-fuelled improved cookstove intervention to prevent household air pollution in Northwest Ethiopia: a cluster randomized controlled trial. Environ Health Prev Med, 26(1). 10.1186/s12199-020-00923-z

Fiks, A. (2015). Parent-Reported Outcomes of a Shared Decision-Making Portal in Asthma: A Practice-Based RCT. Pediatrics, 135(4), e965–e973. 10.1542/peds.2014-3167

Jaakkola, J. (2019). Regular exercise improves asthma control in adults: A randomized controlled trial. Nature, 12088. 10.1038/s41598-019-48484-8

Knibbs, L. (2018). Damp housing, gas stoves, and the burden of childhood asthma in Australia. Med J Aust, 208(7), 299-302. 10.5694/mja17.00469

Seals, B. (2020). Health Effects from Gas Stove Pollution. Rocky Mountain Institute. https://rmi.org/insight/gas-stoves-pollution-health/

[1] RCTs have, however, been performed with other types of cooking equipment. (Adane, 2021)

[2] Based on Google Shopping search for “induction stove” on 12 November 2021.

Originally written for Politics of Policymaking Fall 2021 at Columbia SIPA

Over the last 20 years, a dramatic increase in renewable energy has provided real hope in the fight to prevent the worst impacts of climate change. Along the way, one key success story has been the ability of grid operators to integrate ever-larger amounts of wind and solar power into the grid without affecting reliability of electric service. Yet in February, 2021, when a wave of freezing cold temperatures brought the Texas grid crashing down, anyone watching the news or reading social media would have heard a very different story.

Even as energy experts scrambled for clues to the grid’s implosion, Republicans and allies of the fossil fuel industry lept into action with their own story of what happened. Their story was this: Wind turbines that now supply nearly 20% of Texas power had frozen in the extreme cold, leaving Texans stranded. Tucker Carlson was one of those spouting this story, claiming that “a reckless reliance on windmills is the cause of this disaster.”

As the cold snap continued, politicians and industry flacks quickly pivoted from casting blame to advocating for policy. Texas Governor Greg Abbot told Fox News that the Texas power outage “shows how the Green New Deal would be a deadly deal for the United States of America.” U.S. Senator Steve Daines of Montana tweeted that Texas was the “perfect example” for why we need to continue reliance on fossil fuels - and why Joe Biden’s nominee for Secretary of the Interior was a mistake.

The story was almost entirely inaccurate. There is no evidence that renewable energy has affected grid reliability, and wind power has proven wildly successful in places that routinely see sub-freezing temperatures, like Montana and the Dakotas. Indeed, when the dust finally settled, the problem in Texas turned out not to be an inherent problem with wind power, but rather the failure of state regulators to ensure adequate investment in grid resilience, leading to extensive problems with both fossil fuel and renewable infrastructure.

But despite, or maybe because of its inaccuracy, this is a great example of a causal story. As Deborah Stone writes in “Causal Stories and the Formation of Policy Agendas,” a key function of any causal story is “pushing responsibility onto someone else.” The “blame wind” story did exactly this. As millions of Texans suffered, Republicans saw an opportunity to shift the story from, in Stone’s typology, one with an “accidental cause” to one with actors to blame: “reckless” environmentalists and Democrats. The complexity of and initial confusion over underlying causes made the task of shifting blame even easier.

The story also had key elements that Stone declares as important to success. It had backers with high levels of media access -- including, as noted above, Tucker Carlson, the Texas Governor, and Republican US senators. The story also supported the status quo, and did not entail any “radical redistribution of power or wealth.” Americans depend on having reliable energy, and the causal story was able to portray renewable energy as a source of radical change threatening stability.

For these reasons, the “blame wind” story proved immediately effective. It garnered national and even international media attention, with countless news stories with headlines like “Are frozen wind turbines to blame for power cuts?” These headlines alone _{nuances in story content aside} were a messaging victory for the fossil fuel industry, sparking uncertainty over clean energy. Google Trends shows a dramatic national spike in interest in wind power from February 2021, with four times the number of searches that month as in any other over the last 12 months. The top related topic for wind power? “Freezing.”

In the weeks following the Texas cold snap, news reports began to report the real story of what happened. But major interest had already ended, and it is likely that many Americans continue to believe that wind power _{not the incompetence of Texas regulators} is to blame for the disaster.

This provides a lesson in the power of causal stories, and shows how dangerous they can be when put to nefarious use. In the case of clean energy, the lesson is that a successful transition will won’t just entail deploying new technology. It will also take real work to counter misinformation, educate the public, and aggressively react to energy news stories to ensure that, in any media frenzy, the truth comes out.

Today took a walk to the Harlem River Park.

According to the park website:

The park restores the community’s historic access to the Harlem River waterfront by providing pedestrian access ways over the Harlem River Drive at 135th Street and Madison Avenue, 139th Street and Fifth Avenue, and 142nd Street and Fifth Avenue. Through a partnership with the Harlem River Park Task Force , the park is at the forefront of community engagement and innovative design.

The park offers some nice views of the river, the Bronx and various pieces of infrastructure, including the Harlem River Lift Bridge.

But the park feel pretty underused, likely because it's so hard to get to. The aforementioned pedestrian access at 135th St., for example, is narrow, dark and grim, and today was filled with piles of asphalt and trash that looked like they'd been there a while:

It's nice that New York is working on expanding and improving this park. Riverside parks can be my favorite parts of cities, and in dense, crazy Manhattan it feels especially important to have these outdoor places by the water. Hopefully park plans include not just making the park nicer, but also making it easier to get to.