diff --git a/pypdf/generic/_base.py b/pypdf/generic/_base.py
index 5a2757295..813b1df04 100644
--- a/pypdf/generic/_base.py
+++ b/pypdf/generic/_base.py
@@ -650,4 +650,4 @@ def encode_pdfdocencoding(unicode_string: str) -> bytes:
             raise UnicodeEncodeError(
                 "pdfdocencoding", c, -1, -1, "does not exist in translation table"
             )
-    return retval
+    return bytes(retval)
diff --git a/tests/test_generic.py b/tests/test_generic.py
index 0e0fff677..2a9c5ec00 100644
--- a/tests/test_generic.py
+++ b/tests/test_generic.py
@@ -1,4 +1,5 @@
 """Test the pypdf.generic module."""
+
 from io import BytesIO
 from pathlib import Path
 from unittest.mock import patch
@@ -273,6 +274,16 @@ def test_encode_pdfdocencoding_keyerror():
     assert exc.value.args[0] == "pdfdocencoding"
 
 
+@pytest.mark.parametrize("test_input", ["", "data"])
+def test_encode_pdfdocencoding_returns_bytes(test_input):
+    """
+    Test that encode_pdfdocencoding() always returns bytes because bytearray
+    is duck type compatible with bytes in mypy
+    """
+    out = encode_pdfdocencoding(test_input)
+    assert isinstance(out, bytes)
+
+
 def test_read_object_comment_exception():
     stream = BytesIO(b"% foobar")
     pdf = None