���˴�Ƭ

The Innovators, Entrepreneurs, and Collaborators series profiles regional innovations, startups and collaborations that are flourishing and which engage Queen’s faculty, staff and/or students.

Bahman Kashi slides the mouse in his right hand back and forth on the table top, clicking on the long columns of numbers and graphics displayed on his computer. To the casual observer, seeing these long ranks of columns on a wall screen in a boardroom at Innovation Park, what he is working with looks like a long, complex, and not easily understandable spreadsheet; in fact, these numbers are part of a sophisticated analysis that his company, Limestone Analytics, is carrying out aimed at improving health outcomes in hospitals in Cameroon, specifically saving children’s lives. 

Important, interesting work. But what has earned Kashi and his firm a spot at Innovation Park, Queen’s University’s incubator for startups (usually of the high-tech variety), is what lies behind the numbers – what Kashi refers to as Limestone’s “methodology.”

“Economists love models,” says Kashi, himself a PhD in the field and an adjunct lecturer in the Department of Economics. With good reason. A mixture of data, assumptions and formulae, economic models are powerful tools, useful for determining the costs and benefits of business decisions, government policies or development programs such as the one that Kashi and his colleagues are working on. But the typical model is anything but user-friendly. Experts on a program or economic sector often build extensive models across a series of interconnected spreadsheets, making it nearly impossible for anyone else to update their analysis, let alone understand the details of the calculations being performed. Very much the idiosyncratic product of one mind, “it’s easier,” says Kashi, “for a second person to rebuild it from scratch than understand what had been done.” 

To cap it off, says Kashi, “Economists aren’t the best at communicating.”

Not surprisingly, given the somewhat artisanal fashion in which these models are constructed (they can take literally hundreds of hours), hiring an economic consultant is an expensive proposition. Kashi wondered if there might not be a more efficient and less expensive way to do it, one which would make modeling more accessible to a larger group of potential users. “An architect can draw a plan and pass it to a builder,” he says. Why not do something similar with modeling?

The result is what he is displaying on-screen. It looks like a spreadsheet because it is a spreadsheet. But baked into it, so to speak, are the economist’s assumptions and formulae – relating to costs and benefits, the social impacts, the stakeholders. Different people can work on different parts of the model – a health economist could create the model, essentially a series of equations; a programmer could then incorporate the equations into Excel; and a research assistant (who might be an undergraduate, not a PhD) could track down statistics and enter them in an Excel table that the programmer set up to feed data into the underlying model. Making changes can be done quickly, he says, in a “matter of hours, not a week,” as might be the case with a more conventional model.

Kashi was interested in bridging the gap between the social sector and the world of academic economics. He knew people in the social sector from teaching at Queen’s, where they would approach the department for help with their projects. Largely lacking in economic training, they didn’t have the right conceptual tools to evaluate a program or choose between two competing ones. 

Founded in 2016, Limestone Analytics specializes in economic modeling, as well as the design, monitoring and evaluation of international development and social sector projects. Working with Kashi and his firm gives NGOs (among them World Vision, one of the world’s largest), a sophisticated analysis of potential or ongoing projects. For their part, Limestone Analytics gets real, hands-on examples to help them hone their methodology further. Those real-world examples are key, says Kashi. “We would be making fools of ourselves if we just went into a room for a year and a half, and then came out and said, ‘Here’s the model.’” 

One of Limestone’s recent projects, focused on an analysis of the Haitian electrical sector, undertaken for the Copenhagen Consensus Center, a Danish think tank. Limestone’s project was chosen as the number one submission by an expert panel created by the Center to look at ways to help the Caribbean country climb out of poverty.

Most firms involved in consulting work similar to that done by Limestone Analytics tend to locate in Toronto or Ottawa, or even Washington. Limestone plans to stay in Kingston, at Innovation Park. 

There are, says Kashi, a number of reasons for this. 

“One is academic rigour. Very often these social-sector analyses are critiqued as poor quality, so we want to maintain our relationship with high-quality academic partners in the Queen’s Department of Economics.” 

Thanks to the university connection, they also receive funding from MITACS, which reduces the costs for them to hire graduate students. Other faculty members in the partners in the Queen’s Department of Economics are also regularly involved with Limestone’s projects, helping ensure that they adhere to the highest standards of academic quality.

“The other point is if we were in Washington, say, we’d be flooded with jobs. But you don’t want that if you are trying to change the very way things work. And we wouldn’t get the support we get here,” he says. Now up to eight people, Limestone Analytics has recently moved into a larger, more private footprint within the incubation space at Innovation Park, which continues to provide the company with access to numerous resources such as business advice from Launch Lab, and match-making services and intellectual property guidance from the Queen’s Office of Partnerships and Innovation. (Limestone Analytics can also draw on resources in Toronto and Ottawa.) One of the key proposals Limestone is working on now is a direct outcome of an international event that took place at Innovation Park in June. “Even getting into MaRS (Toronto’s startup incubator) for a desk you’d have to wait a year or two,” he laughs.

“We are not the first to have tried this,” says Kashi of the idea of creating a development-specific economic model. Both the United Nations and the World Bank have tried, but earlier attempts proved unsuccessful. “The problem is that earlier attempts have either tried to create a complex model to fit all situations, or restricted their assumptions to the degree that their work is no-longer useful. Our way is different. We are trying to develop a streamlined approach to the modeling process, while still allowing the models themselves to be flexible in their design and assumptions.” 

Ultimately, their goal is to refine their methodology and scale it up, which will give them a real product that they can sell to the World Bank, or other large and small organizations that deal with investments in infrastructure and social projects. 

“It’s a niche market,” says Kashi, “but there’s lots of demand for it."