How Crisis Pregnancy Centers Spread Across the Carolinas
Mapping three decades of expansion and what it means for reproductive access
In 1990, there were roughly 60 crisis pregnancy centers (CPCs) in North and South Carolina. By 2019, that number had grown to 146. During the same period, the number of abortion providers barely moved, going from 20 in 1990 to 23 in 2019.
This post walks through the data behind my job market paper on CPCs and fertility decisions, starting with the most basic question: where did these centers open, and why does that geography matter?
What are CPCs?
Crisis pregnancy centers are nonprofit organizations, typically faith-based, that offer free pregnancy testing, counseling, and material support (diapers, baby clothes, parenting classes) to pregnant women. Their stated goal is to provide alternatives to abortion. They do not perform or refer for abortions. There are now more than 2,500 nationwide, outnumbering abortion clinics roughly three to one.
Despite their prevalence, there was essentially no causal evidence on whether CPCs actually affect abortion rates before this paper. That gap is what motivated the research.
The expansion, decade by decade
Use the slider below to scrub through time, or hit Play to watch CPCs (pink circles) and abortion providers (blue triangles) appear and disappear across North and South Carolina from 1975 to 2020. Click any marker for details.
A few patterns stand out.
First, the growth is not uniform. CPCs expanded along the I-85 corridor (Charlotte to Raleigh) and in the western mountains of North Carolina. The coastal plain and rural southern South Carolina saw much less penetration.
Second, abortion providers barely moved. The blue triangles are almost identical across all four maps. They sit in the same handful of metro areas (Charlotte, Raleigh-Durham, Greensboro, Asheville, Greenville, Charleston, Columbia). This means most of the variation in reproductive access over this period is coming from the CPC side, not from changes in clinic supply.
Third, by the late 2010s, CPCs outnumber abortion providers roughly 7 to 1 in these two states. Many counties have a CPC but no abortion provider within their borders.
Why geography matters for identification
The key challenge in studying CPCs is endogeneity. CPCs don’t open randomly. They may target areas where abortion rates are already high, or where local religious communities are mobilizing. A naive regression of abortion rates on CPC presence would conflate the effect of CPCs with whatever drove them to locate there.
My paper addresses this with a simulated instrumental variables strategy. I estimate a discrete-time hazard model of CPC entry from predetermined county characteristics, then forward-simulate 1,000 counterfactual expansion paths. Averaging across draws yields a predicted CPC count that converges to the conditional expectation of CPC presence given observable county history. The identifying assumption is predetermination: conditional on rich historical observables and two-way fixed effects, the residual timing of CPC openings is orthogonal to contemporaneous fertility shocks.
The maps above give visual intuition for why the instrument has power. The expansion followed a predictable spatial pattern tied to population density, existing CPC stock, distance to the nearest center, and proximity to evangelical institutional infrastructure. The hazard model captures this path-dependent process, and the forward simulation compounds it over 30 years into cross-county variation that no linear function of the same covariates can replicate (first-stage F = 170).
I go deeper into the methodology in a separate post.
What did all that expansion do?
The headline result: one additional CPC per 10,000 women reduces abortion rates by 13 to 14 percent in 2SLS estimation. Pregnancy rates do not respond to CPC presence, ruling out deterrence (upstream behavioral change that prevents pregnancies) as the primary channel. The accounting is consistent with substitution: CPCs change how pregnancies are resolved, not whether they occur.
But the aggregate number hides important variation by age and marital status. I unpack that in another post.
This post is based on my job market paper, “The Role of Crisis Pregnancy Centers in Fertility Decisions.” You can read the full paper here.
