Join single points to multiple polys

Question

I have a student with a set of geocoded points representing political earmarks. The earmarks are geocoded to the city effected by (or named in) the earmark. I also have a layer of congressional districts with district IDs. The student wants to assign the earmarks and the sum of their values to the congressional districts they fall into. Many of the cities, and therefore the earmarks associated with them, actually fall into multiple congressional districts. The student would like to assign to each district the count of all earmarks in that district and a proportional sum of the value of the earmarks.

My main problem is that the earmarks are geocoded as city points so spatially they fall into only one district, but I need to represent them in each district they actually effect and get a proportional sum on their values (so an earmark in 3 districts would be divided by 3 before being summed into the total earmark value for each district).

DanLee · Answer

Perhaps the following tools can give you some lead (not exactly tested):

1. Run the Intersect tool to intersect the points with districts. One point may intersect multiple district polygons.
2. Run the Frequency tool on the point IDs. This gives you the count of a point in polygons.
3. Run Join Field tool to join the points with the frequency table via Point IDs to transfer the frequency/count values to the point.
4. Use Add Field and Calculate Field tools to get the desired proportional values, by dividing the values by the counts.
5. Run the Spatial Join tool to find the districts that CONTAINS points. You may get the divided values from the points via the Field Mapping.

JoeBorgione · Answer

That's the thing; there's more than one way to skin a cat.....

DavidMedeiros · Answer

A couple approaches come to mind:  the Near command, described here or perform a buffer around each earmark point and then geoprocess the buffers into/with the districts.

I'm purposely leaving this a little vague as I'd like your student to figure out the specifics....

Thanks. We usually just help with general understanding of the software and science itself but with no actual GIS curriculum on campus we often have to delve into holding hands through analysis design as well.

I'd considered buffers but there may be cases where a buffer would cross into nearby districts that don't actually belong to the city the earmark is in. Dialing in the exact right tolerance for a few thousand entrie3s across all 50 states would be difficult if not impossible.

I may have come up with a solution though. I'm intersecting the ESRI placeply layer with the congressional districts to get a layer of polygons that contain both the district id and the city it falls into (or more specifically, pieces of the districts broken into their city portions). Table join the earmarks to the intersected layer by city/sate name then merge the output back into districts only on the district name. Still need to figure out the count and proportional sum issue but I think this moving in the right direction.

Still looking for input though.