In a recent analysis, we looked at the characteristics of the December cohort, which totaled $4.2 billion to 384 nonprofits. At the time, we did not know anything about the grants themselves. Because Scott is not using a foundation, she is not required to release grant-specific data.
Luckily, many of the recipient nonprofits have shared further data. At Candid, we have been able to track down this information using a combination of automated and manual searches.
Our automated tool “reads” about 300,000 media articles each day. We then use machine learning algorithms to identify a subset of each day’s articles (usually around 1,000) that relate to the social sector. For MacKenzie Scott’s grantmaking, we’ve also supplemented our data by manually searching social media and organizational websites.
So far, we’ve identified data on 176 of MacKenzie Scott’s grants, and we’re adding more as we find them. These grants represent about $2.4 billion of Scott’s grantmaking, or almost half of Scott’s grant dollars in 2020—enough to suggest patterns.
Grant size distribution
We identified grants ranging from $750,000 to $126 million. The $126 million grant, to Easterseals, is an outlier in size and structure—it actually represents an aggregation of grants to a set of individual Easterseals chapters.
If we order the remaining grants from largest to smallest, we see a set of clear plateaus where Scott provided grants of the same size. In particular, we see 41 grants of $10,000,00 (23 percent of this set); 21 of $20,000,000 (12 percent); and 19 of $5,000,000 (11 percent). Institutions of higher education—notably historically Black colleges and universities—received the bulk of grants above $30 million.
Visualized this way, this set of grants suggests a more general pattern. Scott made a small number of very large grants and then increasing numbers of individual grants at lower dollar levels. Her grantmaking in 2020 appears to follow a regular pattern of proportionality—perhaps a “power law distribution,” a mathematical pattern found across natural and socioeconomic systems where one quantity varies as a power of another.
Whether or not Scott and her advisors had an actual formula, it seems they intentionally set common levels for many of the grant amounts. This type of approach also highlights a hidden operational advantage to giving general operating support grants. If the funder and recipient do not need to negotiate the exact cost structure of a given project, the funder can achieve efficiency by giving similar amounts to multiple organizations.
Grant size versus organization size
The recipients of MacKenzie Scott’s grants range significantly in size, from modest organizations to billion-dollar institutions. (Further analysis of organizational size can be found in our earlier report.)
Plotting grant size to organizational size reveals a pattern: larger organizations got larger grants. But the relationship is relatively weak (30 percent correlation) and requires a logarithmic scale to see clearly:
Seventy-one of the grants (40 percent) were in amounts that represented half or more of the recipient organization’s latest reported annual revenue. Seventeen (10 percent) were in amounts that exceeded the organization’s latest annual revenue. (This figure excludes Easterseals.)
Grant coding by issue area and beneficiary group
The patterns behind Scott’s grantmaking are much clearer when it comes to issue area and beneficiary group. As she herself stated, her grantmaking was guided by “special attention to those operating in communities facing high projected food insecurity, high measures of racial inequity, high local poverty rates, and low access to philanthropic capital.”
We have no way of knowing how Scott and her team formally categorized either the grantees in the portfolio or individual grants. But Candid’s auto-coding algorithms give a sense of the general categories. It is worth emphasizing that our algorithms will often assign multiple codes to a specific grant. We have purposely set them up to do so, because often a grant will have multiple purposes and serve multiple beneficiaries. Accordingly, the totals below add up to more than the $2.3 billion in grants we’ve specifically identified.
The leading issue category is health, followed by “public safety,” which includes funding focused on disasters
Similarly, the data we see on beneficiary group is entirely consistent with Scott’s recognition that “this pandemic has been a wrecking ball in the lives of Americans already struggling. Economic losses and health outcomes alike have been worse for women, for people of color, and for people living in poverty.”
In particular, we found $881 million directed to Black communities, $60 million to Latinx communities, and $46 million to Native American communities. We’ve identified an additional $480 million in grants directed to serve people with disabilities.
Scott’s 2020 grantmaking is notable for its immense size. But it is also notable for its approach. Scott gave at scale to those with deep need. Her giving was also defined by trust, with grants structured as general operating support, thus handing power from the grantor to the grantee. In addition, Scott made use of another kind of trust: regranting. Of this set of 176 organizations, 30 have been tagged as regrantors in our database. United Ways are the most prominent example, but Urgent Action Funds are also noteworthy—they do rapid response grantmaking internationally in support of women’s human rights.
Scott’s gifts to regranters is good news for smaller nonprofits that did not receive grants directly from her. Regranters typically find smaller organizations to regrant to, so there may still be opportunity for this capital to make its way to other organizations. Equally important, Scott still controls more than $50 billion and intends to, as she has said, keep giving “until the safe is empty.”