For the chosen example of Ellis Island, you are trying to make a distinction where no distinction exists. A passenger list (which might be a passenger arrival or passenger departure list) in this context is a specific type of ship's manifest.
There are other instances where we have passenger lists that are not strictly a ship's manifest (see the US Customs Service records discussed later). Other examples of passenger lists that are not ship's manifests are the lists of arriving passengers published in newspapers (news of Irish immigrants during the Famine period, or lists of survivors of the Titanic). But even these are associated with a specific voyage of a specific ship, and are either derived from the ship's passenger list, or the same source of information as the ship's passenger list -- so I'm not sure what distinction you're trying to make.
The important things to consider for any record are the agency who created the records, and the purpose for which it was created.
The Ellis Island myth
There is a persistent myth that the inspectors at Ellis Island created the passenger arrival lists as people arrived at Ellis Island. Common sense alone tells us that this can't be the case. In her article A Guide to Interpreting Passenger List Annotations, USCIS historian Marian L. Smith gives examples of markings that indicate that the passenger on the list did not sail on that particular voyage. This shows us that the main part of the passenger lists were NOT created at the arrival port.
In her presentation "Fact, Fiction, and Immigration Passenger Lists", which was recently presented as part of the USCIS's "Worth Repeating Webinars" series, Smith showed a copy of one of the intake forms that were given to ticket buyers at the steamship company offices. These were forms that listed the questions that researchers are familiar with from seeing them on the headers of the US passenger arrival lists. Smith showed a blank copy of one of these forms that she found in a file of INS correspondence where people were discussing whether the questions being asked at the departure ports was sufficient for the INS (as the agency was then called) to do their jobs in screening immigrants.
Over time, the number and the nature of the questions changed. A summary of the questions and the time period in which they were in use can be found at the website of the Capital Area Genealogical Society, in their section of ship passenger list extraction forms:
- Ships Passenger List Extract Form [1893-1906]
- Ships Passenger List Extract Form [1907-1918]
- Ships Passenger List Extract Form [1919-1925]
- Ships Passenger List Extract Form [1925-Present]
These four forms created by Lisa Perkins essentially back-engineer the forms that ticket buyers filled out in the steamship offices in the ports of departure.
A much smaller subset of the pages in the Ellis Island arrival lists were created at Ellis Island -- those are the Record of Detained Aliens and the Records of Aliens Held for Special Inquiry which are sometimes called BSI records (BSI = Board of Special Inquiry). These (few) pages were added to the end of the passenger arrival lists.
Finding Passenger lists
An overview of US passenger lists can be found in the FamilySearch Wiki article US Immigration Passenger Arrival Records. Many early (1600s) lists of passengers arriving in the USA were produced from published sources, and these are indexed in publications like Passenger and immigration lists index by Filby and Meyer.
The basic principle to keep in mind is that records are created for a reason -- usually there is a law that mandates the information needs to be collected. You can see the effects of the changes in the law by looking at how the different record sets changed over time. In the US, the Federal Government began keeping track of passengers in 1820. For the port of New York, the early lists from Castle Garden were lost in a fire in 1897. Marian L. Smith researched the fate of the Castle Garden records and presented her findings in 2015 in a presentation at USCIS called "How Castle Garden Records Burned in the Ellis Island Fire". The only surviving records we have about incoming passengers for the Port of New York from this period were created by the US Customs Service. The records from this period have been brought together by the US National Archives as part of Record Group 36 and their online guide to the records can be found here: Passenger Lists: Microfilmed Records of the U.S. Customs Service, 1820--ca. 1891.
A similar situation exists in Canada. Lists from the period before Canadian Confederation have to be assembled from information gleaned from other sources. Olive Tree Genealogy is a site with lots of useful information about these early sources. OTG collects early ships passenger lists and other records which can act as substitutes -- see Poor Law Union Immigrants to Canada 1836-1871 ONLINE for examples of records that can fill in the gaps. After 1865, look to Library and Archives Canada for passenger lists, and see an overview of records from the FamilySearch Wiki article Canada Emigration and Immigration.
For any passenger list - and for any record -- it is useful to find out who holds the original records, and to check that archive for research guides or more information about how to view the records. For the United Kingdom, the two main collections, created by the Board of Trade, can be found online as follows:
- passengers who arrived in the UK between 1878 and 1960 are in the record group BT 26; these are at Ancestry.co.uk
- passengers who left the UK between 1890 and 1960 are in the record group BT 27; these are at findmypast.co.uk
The website The Genealogist has also added records from BT27 to its holdings this week (last week of May 2017) and their feature article about what can be found in those record is here: Ancestral Voyages.
For other countries, check the FamilySearch wiki for the name of the country, plus the phrase Emigration and Immigration.
A good example of how a ship's passenger list is created for a specific agency for a particular purpose is Ancestry's recently-added database, U.S., Army Transport Service, Passenger Lists, 1910-1939. This is a database of two separate record groups at the US National Archives (one inbound, one outbound). Ancestry gives Source information for every database with some kind of note where the original data comes from. For this database, the original sources are:
Lists of Incoming Passengers, 1917-1938. Textual records. 360 Boxes.
NAI: 6234465. Records of the Office of the Quartermaster General,
1774-1985, Record Group 92. The National Archives at College Park,
Lists of Outgoing Passengers, 1917-1938. Textual records. 255 Boxes.
NAI: 6234477. Records of the Office of the Quartermaster General,
1774-1985, Record Group 92. The National Archives at College Park,
Note how NARA always lists the agency who created the records when that is known. Archivists generally provide scope and content documents, reference reports, descriptive pamphlets, finding aids, or other guides to explain who created the records and their archival arrangement.
To sum up: to evaluate any set of records, it's important to understand WHO created the records, WHY they were created, HOW they were created, WHEN and WHERE they were created, and WHAT information you might expect to find there. It's not useful to make blanket statements about any one record group being superior to another without any kind of discussion of what you are using the record group for. The information in a record isn't evidence until we have a question we're trying to answer. See QuickLesson 17: The Evidence Analysis Process Map.