[75% down the page] If the size of firms obeys a power law, economies will comprise some very big firms and a long tail of small ones. The fortunes of the biggest companies might then stir the whole economy, [Xavier Gabaix of New York University] conjectures. The $24 billion dividend paid by Microsoft in December 2004, for example, added 3% to America's personal income that month. Mr Gabaix calls for a more “granular” approach to macroeconomics, which would weigh the contribution of big firms to national aggregates.

This granular view is already taking hold in studies of international trade. Countries, after all, do not trade with each other; companies do. A few firms usually account for the lion's share of a country's exports: in America, the top 10% of exporters account for 96% of the country's foreign sales, and only 4% of firms export at all.

granular = 1. Resembling or consisting of small grains or particles.

2. {technical} Characterized by a high level of [granularity]:

granularity = 1. I omit this circular definition; it just bootlessly refers back to 'granular'.

2. {technical} The scale or level of detail in a set of data.

1. The only possibility seems to be defn. 2 of 'granularity', but this contradicts macroeconomic's already extant, flourishing dependence on data, and this view's novelty as implied by Prof Gabaix's calling for it?

2. What does the bolded relative clause mean? I think that 'weigh to' is confusing me. Are firms' contributions simply being compared to national aggregates?


When a photograph has insufficient granularity, fine detail is not visible. This meaning is often applied figuratively to data, and refers to the fineness of the detail.

If the data are tracked at the state or province level, say, it is impossible to perform an analysis at the city level. The data are not sufficiently granular. A city-level analysis would require data to be tracked city by city.

Granular data can always be aggregated. The reverse is not true: aggregate data cannot be "de-aggregated".


Granular here means level of detail.

I've not read the article and I'm not an economist, but I think what the author is saying is that current macroeconomic approaches deal with the entire economy as a single object. Mr Gaibax believes that in addition to that they should also look at large companies, ie. break down that single object (the economy) into smaller chunks (company A, company B, company C, all the rest) and work with that data and, yes, compare them.


And to "weigh" their contributions means to count them, and consider/examine them, individually; that is, at the company level, not just country totals.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.