Not having an 1890 census as a source is a hindrance for those of us with American ancestors during that time period. While there is a fragment of the census available (and a very small fragment at that), the fact remains that most will never see their relative in the 1890 census.
That can create a research gap which creates challenges. Sometimes writers, advertisers, and the fee-based websites will mention a substitute of sorts. It’s more accurate to refer to anything else (that “substitute”) as a 1890 era source that should be utilized.
City directories, tax lists, voter lists, newspapers are all touted as substitutes. The reality is they are sources that should be utilized anyway, even during those time periods when the census is available. They aren’t necessarily substitutes for the census as they are different records that serve different purposes. The census was intended to count every US resident in 1890–and list them all by name given the time period.
City directories don’t list everyone–generally heads of households and sometimes other adults living in the household who work outside the home. Tax lists in the 1890 era list individuals who had to pay a property or personal tax (for those areas that had them). Voter lists only list those who were registered to vote (and in 1890 that was adult male citizens over the age of twenty-one). The goal of a newspaper is not to name or list every citizen either, although in some rural areas they occasionally mention more people that one thinks.
There is no 1890 census substitute. There are just other records, with different purposes, that may help to fill in some of the gaps.